ACT A I' ,.\ L ,\ a 0 N T 0 ". L ,0 G I C A I' 0 LON IC . ,\

Vol . II 1957 No.4

JULIAN KULCZYCKI

UPPER FROM THE HOLY CROSS MOUNTAINS (POLAND)

A bstract. - The present paper deals w ith fossil r emains (, Elasmo­ branchii) from the Upper Devonian of the Holy Cross Mountains. The following new fo'rms have been described: M al ero st eus gorizdroae n. gen., n. sp.; T omaiosteus flrossi n. gen., n. sp .: denisoni n. sp.; D. ceterus n. sp.; ko ­ ilowsk ii n. sp., D eveonema obrucevi n. gen., n. sp.; Operchallosteus vialowi n. g an .. n. sp. ; AUenacanthus malkowskii n . gen., n. sp .; Sentacanthus zelichowsk ae n . gen., n. sp . The presence of D inichthys pusrutosus, D. cf. tuberculatus, Pachyosteus butia . H %nema radiatum, Anomalichthys ingens , Plourdosteus sp ., Oxyosteus sp., St eti­ ost eus? sp., Ctenacanthus sp. , as w ell as of som e detached teeth of Clado dus a nd Dittodus, has been recorded. On the base of the in vestigated material the a uthor agrees with the op in ion s postula ting th at in Bra ch ythoraci there has occured the disappearance ot dentine and its substitution by oss eous tissue. In the brachytho ­ racids this process had taken place at a considerably ear lier evolutionary stage th an in th e Crossopterygii and the Dipnoi and it had. moreover. progress ed farther, having affected the jaw denticles to o. The structure of the parasphenoid in Pachyosteus, problems relating to the position of the gill slit, and ch anges in the outline of bones during ontogeny are discussed and some cursory remarks on the systematics of are given. The final chapter contains notes on the stratigraphic distribution and geogr ap h ical range of the described c ra chyth oracid forms.

INTRODUCTION

This paper is concerned with .placodenm remains from and deposits in the Holy Cross Mountains. In accordance with information published by J. Czamocki (1947), tW:J sedim en tation zones may be differentiated within the region of the Hol y Cross Mountains. Their boundary line runs approx imately conformably to the axis of the Kie1ce syncline. The northern Lysa Gora region differs from the sou th ern Kielce region in more complete development' of th e paleozoic series, in facial differences and in considerably greater thick­ ness of sediments which are frequently of Flysch character. All of the Upper Devonian fish material has been recovered from the Kielce region, namely from the localities of Wietrznia, Psie GOI1ki an d Kadzielnia, and from the village of Galezice lying south-west of Kielce. Nco Cl> ImmlHI FomennlOYl

~ FrOSnIOl"l ® Ps ie ·Gorki f"7777] lower a. MIddle ~ I- rLLL.aDevonion \.,.../ w.etrzruo I'77'"""A Predev~n ia J'1 @ Cctezrce r.L..L.J deposirs

Rutka ----. c.. ~ ~ t>: sc:: ~ o :>1 H

Fig. 1. - Distribution of localities of fis h re m ains fi nds. Geolog y simplified. after a m a p by J . Czarne cki. Scale 1 : 100000. UPPER DEVONIAN l"ISHES 28 7

In the adjacent hills of Wietrznia and of Psie G6rki the Frasnian beds are similarly developed. They are represented here by layered limestones and thin layers of shales. The follow ing three horizons have been diffe­ rentiated by Czarnocki (1947 and 1950; see also Rozkowska, 1953) on the recorded brachiopod fauna: Upper Frasnian -" horizon III with Hypothyridina cuboides and Man­ ticoceras intum escens, Middle Frasnian - horizon II with Hypot hyridina coronula, Lower Frasnian - horizon I with Hy poth yridina procubo ides. The specimens of fishes h ave 'been recovered Irorn all Frasnian hor i­ zons iJn the Wietrznia hill. In Psie G6r1ki the fishes h ave b een yield ed by horizon III only. In the Kadzielnia hjll are d eveloped th e three Frasnian horizons and a considerable part of th e Famennian. The Lower and Middle Frasnian beds consist here of reef lim estones.F ish remains have been found in Upper Frasnian and Famennian beds only. The Upper Frasnian here is developed as thick bedded limestones w ith Hypothyridina cubouie s and Man ticoceras in tum escens. The Lower and Upper Frasnian consist of shales and m arls with a fauna of cephalopods represented by the genera Cheiloceras and Tornoceras. In the fourth locality situated at the village of Galezice (about 15 km to the south-west of Kielce) there are no Frasnian or Lower Famennian beds. Upper Famennian beds, reduced to a thickness of 3 to 4 m, rest di­ rectly on the Givetian . They are represented by limestones from horizons containing Prolobites, Platyclym enia, Orthoclym enia, Goniocl ymenia, and Wocklumeria, the latter underlying . d eposits. ... The material descr ibed in the present paper is composed of two different collections: 1) the specimens collected b y the 'late J. Czarnecki before the w ar and par tly d estroyed through war activities. This collection is housed in the Geological Institute (I. G.) in Warsaw; 2) the specimens collected b y ·th e lat e P rof. Z.Gor izdro-Kulczyoka and by the p resent author b etween th e years 1947 to ).952 on beh alf of the Muzeum Zi emi (Museum of t he Eae th) in Warsaw. Work at further enla rgement of the collection was con tinued by the w riter during 1953 to 1955 on behalf of the Paleozoological Ins titute of the P olish Ac ad em y of Sciences in Warsaw. This collection is housed in the Mus eum of the Earth (IVI. Z.) in Warsaw. The material was mostly p repared by treatment in acetic acid and .satu ration in paraffin. Before being photographed the specimens were whitened with ammonium chloride. 288 JULIAN K ULCZY CKI

The labor atory work has been carried out in th e years 1953- 1956 ,at the P al eozoological Institute of the P olish Academy of Sciences in Warsaw . Throughout the investigation m ost helpful advice and assistan ce were given to the w riter b y the H ead of the ab ove m en tioned Ins titute, Prof. Roman K ozlow ski. In the course of the f ield work the m ost fr iendly kindness and h e1p w ere sh ow n to the w ri ter by the Managers of the Museu m of the Earth - Prof. 81. Malkowski and P rof . A. H a'lickD. Dr. M. Zeli chowska, ch ief of th e Dep artment of D ocumentat ion o f th e Geolog ical Ins titu te in Wairsaw , has provid ed the autho r witlh the OIPiPOII'­ tunity for studying t!he specimens in h er charge. In 1955 and 1956 the au tho r spen t two mon ths at the Geol.-Pal eonto­ logical Institute of the Humboldt Un iversity in Berlin. Prof. W. Gross h as prov ided the au thor with the Iaci lities for s tud y in g at this In st itute and offered very useful criticis m. Much valuabl e in form ation and advi­ ce w as obtained by the writer, in the way of correspondance, from P rof. D. V. Obruchev of Moscow and Dr R. H. Denison of Chicag o. Mi ss A. Illner , the libr arian of the Paleontological Institute of the Hum­ boldt University in Berlin, has taken pains in supply in g t he au thor with th e required literature. The ac com panying p hotographs have b een taken by Miss M. Czarnocka . The paper has been transl ated by Mrs. J. Hum­ n icka. To all these p erson s the w riter w ishes to expr ess h is deep es t gra­ titude.

DES C RIP T I ONS · Order Infraor der Br ach ythoraci­ Genus Pl'ourd osteus 0rvig, 1951 Plourdosteus sp . (pI. I, fig.I , 2)

MateriaL - A fragmentary p osterior ventro late ral and a comp lete marginal (pI. I, fig. 1). T hese t wo elemen ts were fo und lying beside one another on a sm all rock fr agment and d isplay a sim ilar ornamentatio n pattern suggesting their appurte nance to th e same individ u al. Description . - P osterior ventrolateral relatively larg e, the p rese rved frag ment bei ng 38 mm long and 28 mm wide . Its an ter ior lateral por tion shows the base of a br ok en off process, w hich in this genus is commonly connected with the p oster ior l ateral. The ou ter surface dens el y cover ed ,by minute tubercles , som ew hat larger al on g the m argin, more so in the: posterior p ortion. The associated marginal is markedly s m all. Its shape (pl. I, fig . 1) indicates th at it was not united with the central. th is being UPPER DEVONIAN FISHES 289 a characteristic feature of genus Plourtiosteus. The course of su ture wi th ' th e postorbital and the paranuchal show the same in de ntures as in Plourdosteus tr autscholdi (Eas tma n) (Obruceva, 1954, fig . 1), except for the side edge being more arched so that on the w hole the marginal is more quadrangular . Th e width of this ele ment measured from the poin t of juncture with the postorbital and the paranuchal is 15 mm. The same is its maximum length measured normally to the line. ~ f the p revious dimens ion. The sensory line groove transversing the m argin al produces a rather sharp bend from w hich branches off a blind ramification. As compa r ed with Plourdosteus trau is chol di this rami fication is placed mo re an te r iorly and has a m ore distinc tly la te r al direction . Judging from the shape of the m arginal th e h ead sh ield m ay be supposed to h av e been markedly short, with its posterior side corners exte nding far to the sides. Th e lack of a margin join ing the cen tral refers this specim en to th e genus Plourdosteus. In the same bed as the a bove mentioned sp ecimens was also found th e left infragnathal (pl. I, fig . 2) which, perhaps, may have belonged to the same form. Its entire length is 34 mm, of which 16 mm ' are assign ab le to the functional an ter ior portion. The pos ter ior "blade" is fairl y broad, attaining centrally a width of 8 mm. The upper margin nearly rectilinear, while the lower one is gently ar cua te. On the outer su rfa ce th e lower and th e posterior po r tions are clearly separated from th e smoother, upper-front portion; the latt er passes in to the an te rior functional por tion of the infragnathal. Dur ing the life -time this h ad prob ab ly been covered by the mu cous lin ing only. Both the low er and upper margins of the functional portion are damaged, but it seems doub tless that the latt er was p rovi ded with a n umber of den ticles. Tw o of these are preser ved within th e symphyseal part. Two such complete denticles are also observab le at the back of the functional portion. They are con ical an d 2 to 3 mm in h eight. At the base of ea ch of them there is a slight con vexity on the ou tside of the jaw. These con vexities reveal that: in the functional portion there were five such den ticles over a di ­ stan ce of 7 mm. Occurrenc e. - The Frasnian, horizon II , Wietrznia hill in Ki elce.

Br achythoraci gen. et sp. indet. (a) (text-fig. 2; pl. I, fig. 3 a-b)

Mate1·ial. - A median dorsal (pl. I, fig. 3 a-b), in a fairly satisfactory state of preservation, has been recovered from the same beds as the above described specimens. Description. - It is elongate, being 41 mm long and about 28 mm wide. Transversally it is rather strongly arched. The maximum width is

Act a, Palaeontologlca Polontca - Vol. II J4 l ~ 290 JULIAN KULCZYCKI

atta ined at the anterior, arcuate margin. The side margins ar e first gently con cave and th en convex, running oblique ly posteriorly towards the m edian part to converge at a nearly r ight angle. The outer surface is densely covered by minute tu bercles arranged on the circumference in concen tric rows. On the ventral surface th e h in d portion of th is element is limited fr om the front p ortion by an arched elevation running brans­ versally at a distance of 8 mm fr om th e hind end of the median dorsal . Me dially this elevation constitutes th e ro unded base of the cari nal pro­ cess. This base continues anteriorly into the cr est running over a distance of 13 mm. The described median dorsal with a broken off crest may be taken for this element of Cozcosteus F ig. 2. - Brachythoraci or of Plourdost eus but lacking the hind spine­ gen. et sp . indet, (n): longitudinal section of -like process , as is the case in jletti the m edian dorsal w ith (Watson, 1932). However, the cr est (fig. 2) d esc ernible crest outlina. preserved in the cast ascends ver y gently to the hind, in opposition to what we observe in these gen era. It attains its maximum height of 6 mm very near the base of the carinal process. The latter descends at an angle of ca. 70° in relation to the median dorsal blade and basally shows a sp oon -like imp ression . Thus, in what the sh ape of the carinal process and of the crest is concerned, our specimen is more like genus Pholidosteus, differing from it in smaller height of the crest and in a terminally pointed hind end of the crest. Hence, the carinal process no t only d id no t protrude beyond the posterior margin of the element, as is th e case in Pholidosteus, but was de eply concealed un de r it. At th e present mom en t: the systematic position of the studied median dorsal may be only tentatively established, although it is !possibl e that we are dealing here w ith a n ew form. Occurren ce. - The Frasnian. horizon II , Wietrzn ia hill in Kielce.

Genus Mal eros teus I '11. gen. Diagnosis. - Brachythor aci w ith h ead shield broad, w ith r ounded side margins separating th e descending lateral portions from th e flattened vault . The large orbital cavity is bounded by the subor bital, preorbital and postorbital. The structure of the cheelk p art is as in Ph olidost eu s or Brachyosteus. Postpineal opening lacking . Th e jaw apparatus is of th e crushing type. The infragnathal has a moderately large crush ing surface of the upper margin and a tooth-like process at the fron t of the func­ tional portion. The posterior supragnathal is broad, with a lateral process

t The generic name is an allus ion to the crushin g jaw app aratus, being deri­ ved from the Greek p.aJ' Epoa, m eaning crushing. UPPER DEVONIAN F ISH ES and a well d ev elop ed crushing surface provided wi'th twotubercles: one placed more or less medially, the oth er at the top of the later al process. The anterior supragnathal of the catching type attached to a too th-like process behind w hich m esiall y the re is an depression toaccomodate the tooth-like infragnathal process. Rem arks. - This genus is monotypic, e rected to include Malerosteu3 gorizdroae n. sp. In the disposition of the skull bones this Iorrn resemble s Pholidosteus from which it differs in the stronger development of the crushing jaw apparatus bringing it closer to Mylostomidae and still mQ;I'~ so to Dinomylostoma. From the latter genus, however, it differs in the less developed infragnathal crushing surface, a lower and more conspi­ cuously, li mited hind portion of this element and in d iff er ent p attern and de gree of development of the posterior supragnathal tubercles..

Mal erosteus gorizdroae 2 n. sp. (text-fig. 3-6; pl. I , fig . 4- 7; pI. II; pl. III, fig. 1-3),

Material. - .A large skeletal fragment of one ind ivid ual, found in the Psie Gorki hill. It comprises t he preorbitals, suborbital, marginal, paranuchal, infragnathal, posterior supragnathal, anterior supragnathal, parasphenoid as well as fragments of 'th e dermal and endocran ial bones , The suborbital and postsuborbital found in the Wietrznia hill are refe­ rable 10 the same species. Holotype. - Incomplete head shield and the anterior dorsolateral of one in dividu al (M. Z.) (pl. I, fig. 4-7; pl . II). Diagnosis. - Mal erosteus gorizdroae is thus farttie on ly k nown spe­ cies of this genus so that the specific and generic diagnosis are the same. We m ay add h ere th at' our species is characterized by an ornamentation pattern consisting of round ed tuber cles of various s ize, the whole rather large, p articularly so in the hind part of t'he h ead . These tubercles are arr an ge d densely and ir regularly. Description. - Th e preserved !parts indicate that they bel ong to a broad-headed form, with a blunt snout and distinct ly m arked side edges ("Dorso-later alkan ten " of Gross, 1932), produced by t he angular dis position of the lateral and do rsal parts of t he skull. At the front 'of the orbital fenestra this angle is about J. 20", while in the portion cor res­ pondi ng to the paranuc hal it is as much as 130 to 140°. From a side view this skull rese mbles t hat of Pholidosteus friedeli J aekel as reconstru cted by Jaekel (1907) or by Gross (1932). The similarit y also concerns the position and relative dimensions of tihe orbital fenestra with a d iameter

! This species is named in honour of the late Prof. Z. Gorizdro-Kulczycka w ho initiated research w ork on the Devonian fishes from the Holy Cross Mountains. 292 JULIAN KULCZYCKI of 60 mm in the specimen here considered. Similarly as in Pholuiosteus and the majority of brachythor acids from Wildun gen the m arginal here ex te n ds into the ch eek area , while the su borbital and the postsuborbital are n ot fr ee, t h is being a feature ch ar ac teristic for example of Dinichthys and Coccosteu s. The entire outer surface is covered by rel atively larg e'. roundish tubercles varying in size from 0.5 to 2 mm, on the who le larger in the posterior part of the head. Where the superficial layer of the , bon e h as been worn off, t'he tubercles are considerab ly smaller, sugges­ ting that their size was dep end en t on dimensions and ag e of the in divi­ dual. The central is preserved as a rather small fragment of the anterior process united with the m esial margin of the right preorbital, This pro­ cess was less slende r than that in Ph olidosteus fri edeli Jaekel. The pineal h as been preserved as fragmen ts lying quite close to the anter ior process of the central indicating that th ese two bon es touch ed ea ch other. The pineal foramen is open. A sh allow groove is indicated at on e side of it only, this asymme try suggests the presence of another pineal foramen , as is the case in Pholidosteus friede l( J aekel (Stensio. 1934a). The p reor bi tals (IPI. II , fig. 1) are broad (70 mm), their length is one and a h alf times as 'large as the width, being over 100 mm. The mesi al margins have been d estroyed but the fad of th e .preservation of th e p ineal and the central suggests that the preorbitals of both sides were not in contact . The posterior por tions of these bones dorsally overlapped the centrals and the postorbitals, The anter ior part of the p reorbitals forms the anterior margin of the orb ital cavity. It is p rovided with a not ch. From the top and inside of the orbital cavity a groove runs downwards onto the ou ter sur face terminating within the anteri or angle of the orbit al fenestra. From the end of that groove, another, less di;­ st inc tly marked , wi th a number of foramina, r uns ve rtically, and ex­ tends to the supraor bital line ("preorbital", according to the terminology of Heintz, 1932), without, however, reaching it. The supraorbital line is indicate d ca. 5 mm abo ve the eve n tual point of juncture with the prolon­ ga tion of the mentioned groove. The absence of traces of the supraorbital lin e below that sector cannot be ascertained quite doubtlessly owing to the bad state of preservation of that portion of the surface. In the anterior part of the preorbital, the supraorbital line runs clos er to the margin of the orbital fenestra than it is the case in Pholidosteus friedeli. On the ventral side a crest is visible on the preorbital, corresponding to the boundary of the "lateral consolidated part" (Heintz, 1932), se parating the median part which coats the endocranium from the lateral part .which constitutes the vault of the orbit. UPPER .DEVONIAN FISHES 293

The suborbital (pl. II , fig. 1, 2) is p rovided with a h igh posterior blade (over 55 mm) whose preserved portion constitutes the margin of the orbital fenestra and w ith a slender anteri or part, the " handle" of Heintz. In this "handle" we may distinguish the upper ornamented par t and the lower list constituting what Hein tz (1932) calls the "to ngue shaped part of handle". According to Stensi6 (1934a) it is connected by its lower ed ge with the palatoquadrate. As is the case in other Brachythoraci, bot h these parts are se parate d here too by an incision which S tensio (1934a) call s "inc isura maxillaris buccalis". From the mesial side of the an terior par t of the su borbit al runs a broad horizontal list limiting th e or bit at the bas e. The infraorbital canal runs on the outside with the fore-part of the supramax illar line arising somewhat above the point' of ju nction of the postorbital and subor bital par ts. This branching is particularly conspicuous on the specimen of the subor bital from Wietrznia (pI. II , fig. 2). The postsuborbital (pI. III, fig . 1) recorded from th e same beds of the Wietrznia 'hill b elongs to anoth er individual, p robably of larger dimen­ sions. In the antero-posterior direction and normally to it the postsuborbi­ tal is curved, it's . con vexity being directed outwards. This is associate d w ith the box-like shape of the sh ield characteristic of the here studied Iorrn, though to a smaller degree than it is of Ph olidosteus. At the base of the an ter ior part, the p ostsuborbital is p rovi ded wi th a p rocess whose margin is overlapped by the su borbital. The anterior a:nd upper marg ins limit a proj ection which most likely p enetrate betw een the su borbita l an d the postorbital. The 'length of this element m easured a'long its free margin is 90 mm. Measured normally to it the h eight of the oute r ornamen te d s urface is 55 mm. The marginal (fig. 3; p l. II , fig. 4) is p reserved on the right side only. From behind and t'op it bo unds the paranuchal, from the front th e posto rb ital . It 'is elongated , wi th the longer axis somewhat ob liquely tr aversing the head sh ield. Its low er end considerably approaches the posterior mar gin of the skull. Nev erthel ess, since this p or tion has not been preserved, it cannot be ascertain ed whether the margin al reached to the free margin of the head shield or was separ ate d from it by the postmarginal, as

Fig. J. - Mal er ost eus gori zdroa!? n. gen., n . sp.: diagr ammatical r econstruction of skeleton, side view; missing parts dotted. ADL anterior d orsolateral, IG i n­ fragnathal, M ma rg -na l, PN pa ra nu ch al , PrO preorbital. PSO postsuborbital. likely connected w ith tlhe postorbital thus separating the m arginal from the central. Similarly as in Jaeke l's r econstruction of Pholidosteus there are no traces here of the posterior portion of the occipital line (called so a fte r the nomenclature of Gross, but by Stensio referred to as the "po-

I P9

F ig . 4. - Malerosteus gori zdroae n . gen., n. sp.: post erior m argin of para nuch al; to joint socke t, pg jo int process, x lateral edge.

sterior p it-line"), whose (presence in Pholidosteus has, however, been mentioned by Gross (1932 b, p. 14). The por tion forming a sheath for the postero-lateral ends of the nuchal (Heintz's "hind thickening") is not so conspicuously arched ventrally as may be observed in Dinichthys. In consequence, the p osterior lateral impression, as it is called by Heintz; is not so clearly limited. At the sam e time, in Malerosteus we can distinguish that: part which forms the joint t hickening. The joint socket U PP ER DEVO NI AN l"ISHES i~ relatively longer (31 mm) and narrower, and d isplays greater uniformity of diameter than in Dinichthys, being here 7 mm laterally and 6 mm mesially. The low er and upper " joint processes", as called by Hein tz, are not d eveloped . The p osi tion of the top of the upp er process is m arked by a r ib reach ing h ere to the margin of th e socket. The glenoidal p ro cess w as short (15 mm as m easur ed outs ide at the m argin of the ornamented paranuchal sur face) and obliquely pl aced . I t!s outli ne does not protrude b eyond the margin of the joint socket. The infragnathal (p l. I, fig. 7 a-b)i's fr- agmentary , but' with the functional part preserved on the left side. Its ante rior symphyseal end is provided w ith a rather large tooth-like process, 10 mm high. Dorsally, beh in d it stretches the crushing surface, which is mo de rately broad (E mm fron tally and narrowing posteriorly). It has a mesial impression to ac commodate a tubercle presen t on the posterior supragnathal. The en tire functional portion of the in fragnathal is an almost perfectly flat, unifor mly thick plate. At its lower m argin there is a g roove for the mandibular line, running antero-rnedially. Its presence indicates that Meckel's cartilage w as not in contact with the lower margin of the infragnathal, as has been figured in Stensio's (1934) reconstruction of Leiosteus, but that it occupied m ore m es ial position in r elation to th e infragnathal, The len gth of the functional por tion of the infragnathal in Malerosteus is 62 mrn, the height being 24 mm. The posterior supragnathal (fig. 5; p l. I, fig . 6 a-c) is an elongated plate narrowing posteriorly. Frontally it is provided with a lateral process, as in Coccosteus, b ut without the vertical row of dentides. The lower side has a fairly broad surface, convex throug hout, which, corresp ond ing to the forrn of th is element, is o-homboidal and has a triangular excavation bounded by two diagonals. The exte rior corner corresponding to the la teral process is provided with a m inute tubercle. A larger one is sit uated mor e or less centrally, i.e. nearer to the mesial angle. The anter ior mesial surface is concav e, both vertically and h ori­ zontall y ag reeing with the shape of the con tiguous surface of the an te r ior supr agnathal, The length of the posterior supragnathal is 48 mm, the width 32 mm, when measured in the upper 'Part of the element' from the apex of th e upper process, wh ich is h ere almo st h or izontally pl aced, to the side m argin of the lateral process. 'Dhe frontal- wid th of the crushing surface is 16 mm; maximum height, together with the tubercles, is 18 mm , of which 2 mm are involved by the mentioned tubercle. The anterior supragnathal (fig. 5; pl. I, fig. 4 a-b, 5) of b oth sides have . been preserved, though damaged. Th ey are in the shape of ho rizontally curving plates, Along the curving edge is a rounded rc rest elonga ting downw ards into a strongly tooth-li.k~ !process ] 0 .mm in len gth. 291; JULIANK ULCZYCKI

Mesially, behind the mentioned process, there is an excavation to fit a similar process of the lower jaw. Farther to the rear stretches a small flat area which is- th e extensi on of the crushing surface in the posterior supragnathal. The upp er p ortion and the anterior margin are damaged in both specimens. The h eight of the anterior supragnathal cannot have been m uch in excess of 22 mm (not comprising the tooth-like process). The maximum attained thickn ess is 7 mm.

ASG ,

\ \ PSG \ -, , - -- B

\ -,

-.~ i.p.

mi:: A ~ ~ ,

~ .

. Fig. 5. - M aler osteus gorizdroae n. gen., n. sp. ; jaw appara tus: A upper eleme n ts, ventral view, B same, side view, C jaw apparat us, medial vi ew , D":E posterior supragna tha ls, ventral view, D Dinomylostoma beech eri, E Mylostoma v ariabile, ASG a nter ior supragna thal, IG inaragnathal, PSG posteri or s upragna tha l, p I lateral process, posterior supragn athal , mt middle tubercle. U P PER DEVONIAN l"ISHES 297

Th e parasphenoid belonging to this specimen has already been individual ly described in a [previous no te of th e w riter (Kulczyck i, 1956) ,as an element of Br achythoraci n . gen . The anterior dorsolateral (pI. II , fig . 3) is here incomplete, as mos t of the bones of this specimen.The se nso ry line runs on ' the outside, reach ing to the hind margin of the bone, Iperhaps even beyond it. Th e course of the lat eral line aiso indi ca tes the d or solateral m arg in, along wh ich the lower an d the upper parts of the anterior dorsolateral lean against one another. The angle thus produced by them nea r the an terior margin has 140°, while posteri orly it approaches' 180° and the dorsol ateral e dge disapp ea rs. On the m edian side, along the poste rior m argin , a fl a t area is indicated, ov erlapped by the pos terior d orsolat eral. On th e anterior damaged edge there is an in conspic uous thickening suggesting th at the con dyl e was no t develop ed very strong ly. The length of the pre­ se rved pa r t is 85 mm. The thic kness at the anterior margin is up to 18 mm. As is sh own by its description , the head shield of Ma le ro s ten~ displays m ark ed r esemblance to that of Pho luiosteus. It is to th is genus th at Gorizdro-Ku1czyck a (1950) has referred Iragrnents of head shield from Czarnecki's collection. Never th eless, the evidence acquired on a mor e complete spe cimen, particularly that fro m th e chara cteristic jaw app aratus, calls for the assignment of this Ionm to a new genus . A spe cimen of the posterior median ventral bon e (fig. 6; pl. III, fig. 3) in a satisfa ctory sta te of !p,reservation , found in th e Frasnian of the Wietrzn ia hill, can only ten ta tively b e assigned to the same species . Its tubercul at ion p attern does no t excl~d e the p ossibility t.hat this eleme n t is r eferable to genus Malerosteus whose range should in this case b e shifted to th e Middle F rasnian. Specimens of M. qoriztlroae from the Upper Frasnian , however, do no t exhibit a conc entric arr ange­ ment of tubercles, which is here qu ite conspi cuous near the ossificatio n centre. The IPost'er ior median ventral here consider ed (fig. 6; pl. Ill, fig. 2) is in the form of a large plate, 107 mm long and about 85 mm broad. It is gen tly tr ansversely curved in dica ting that the ven tr al shield of thi s spe cies was slig htl y con vex. Anteriorly there is a small area, 12 mm in length and 32 mrn in width, to fit the anteri or median ventral. Fuirly large areas stretch along the side edges to be overlapped by the ventrolaterals. Further to the fr ont are somew hat r aised areas , a bout 60 mm in length on every side an d about 8~ mm in width, overlapped by the anterior ven trolater al. At the back a lowered and smaller area is observable to fit the poste rior ventrolateral. Its maximum width is 20 mm while the length is 58 mm. The free area of the posterior median ventral, covered by tuberculation, has th e outline of a cross with a blunt for e- 298 JULIAN KULCZYCKI arm, while the others are p oin te d. Its tu berculation resembles that characteri stic of M. gorizdroae. The element described above is in teresting since we can observe in the middle of its free surface an area d istinctly limited posteriorly and ex hibiting a fin er tubercula tion with a radial arrangement. This area most likely corresponds to the juvenile stage of the posterior median. ve ntral plate. Its outline indicates ...... ~ that considerable changes in the shape of this elemen t must have occurred -,-. along with its growth. In a young individual it w as shor t, its ou tli ne being an almos t regula r pentagon -, with gen tly concave margins. The i differences occurring between a ju­ venile and an adult specimen are so great as to mak e possible the assign­ ment of the ir deta ch ed skeletal ele­ ments to diff eren t taxonomic units. A nuchal (pl. III, fig. 2 a-b ) in ...... a fairly satisfactory state of preser­ va tion has been found in the same .... beds as the specimen described above . It has a similar ornamentation, con ­ sisting of coarse tubercles whose arrangement is concentric around the Fig. 6. Malerosteus goTi zdroae? , poste­ ossification centre only where they r ior med ian ve ntral. Thick continuous are of smaller size. Ventrally, near line - outline of free area in ad ult individual ; thin continuous and br oken to the hind margin a double socket line - outline of fr ee area in young distinct ly divided centrally is discer­ individual; dotted line - outline of the who le eleme nt in adult individ ual. nible. The par tition that divides it is broad anteriorly, narrowing down to the back. In addition a relatively low transversal ridge occurs here, behind which on the hind edge the re is a mi nute central process. The preser ved fr agment is 40 mm in length alon g the central line and 73 mm broad along the hind margin. Occurrence. - The Frasn ian , hori zons II ? and III in the Psie Gorki and Wietrznia hills. Genus Tomaiosteus» n. gen. Diagn osis. - Brachythoraci, with the he ad broad and flattened, at least in the an terior part, and mo derately large orbits placed near the

S The generic name from the Greek word tOf-la toa - "cut ting" is an allu­ sion to the cutting type of the jaw apparatus. UPPER DEVONIAN FISHES 299 anterior end of the skull. The pineal touches the centrals and the rostral, thus completely separating both preorbitals. The jaw apparatus of the cutting type not too highly specialized, Posterior gnathal with preserved lateral process. Remarks. - This genus is monotypic, erected to include Tomaiosteus grossi n. Sip. From other forms with a cutting jaw apparatus this genus differs in the presence on the posteeior supragnathal of the lateral process which is here, however, not so strong as an Coccosteus and similar forms. From the latter, Tomaiosteus also differs in a poorly developed upper process and in general outnine of the hind upper jaw element.

Tomaiosteus grossi n. SIp.4 (text-fig. 7; pl. III, fig. 4, 5). Material. - The preserved fragment represents an almost complete pineal, a major part of the preorbital, small fragments of the centrals and a satisfactorily preserved posterior supragnathal. An indeterminate endocranial fragment was also found together with these remains. Diagnosis. - The same as for the genus, but also sculpture characteri­ stics, consisting of irregularly disposed tubercles of varying size and with diameter of 1 mm or less. Holotype. - Fragment of fore....part of skull with tJherposterior supragnathal (1. G.) (pl. III, fig. 4, 5). Description. - Pineal (fig. 7; pl. III, fig. 4) is an elongated bone whose length measured medially, without the portion missing in the specimen but together with the po­ sterior process, is 40 mm. Dorsal, tuberculated surface 15 mm in length and minimum 7 mm in width. The pineal plate overlaps the centrals wedging in between their dorsal surfaces. In this way, when seen from the ventral side of the head shield, this process appears as a triangular element, half its actual length. The Fig. 7. - Tomaiosteus grossi n . gen., remaining free margins preserved in n. sp. ; the pineal: A ventral view, B dorsal view. the specimen, have grooves fitting the corresponding ridges of the mesial preorbital margins. At a distance of 25 mm from the apex of the medial posterior process is the minute

'( This. species is dedicated to Prof. Dr. W. Gross from the Geol.-Paleontolo­ gical Institute of the' Humboldt University in Berlin. ,300 JULIAN K U LCZYCKI pineal foramen opening on the ventral surface of the 'bone in to the ipiriea l :depressi on. Th is hollow is sh arply limited by a d istinc t: ri dg e while.to the .Iron t it gradually passes in ter the ven tral surface of [he pineal plate. It lis no tewor th y that the pineal foram en is placed somewhat la terally to th e plane of sy m metry indicate d at the bo ttom of the p ineal foramen by a slight' elongated ri b. Th is is also a sign that the pineal opening wa s origin ally paired. Th e preorbital (pl. III, fig. 4) m ust hav e been very broad and almost fla t. It s median margin is arcuate, at first adjacent to the rostral. The p r eorbital was connect ed with tlhe la tter by a n arrow bu t rather long , transversally di rected embayme n t in the anterior par t of th e median margin. Th e orbita l m argin or t he preorbital is not' p r eserved. The lim it. however, of the orbital vau lt, vis ible from the ventral side, where we can also see a fr agment of th e postorbital, suggests th a t the or bits were relatively small, most likely much less than 45 mm in diameter. In the anterior m edi an portion of the orb ita l vault is a funnel-like impression. Abo ve lit in th e crest centrally lirnining the va ultin g, a n otch amd a groov e occur. probably serving as a duct for a nerve or vessel branching . The preserved fragment of th e central indicates that this element was more than 60 rnm broad in the an terio r par t. ThE ante rior m argin of its lateral por-tion lis n early straigh t, transve rsal to the cr anial ax is.

:.vre5ially it d evelops a broad triangular process, abo u t 20 mm high and (1,'; much basally broad. 'I'his process fits between the preorbital and the posterior pmeal process . The posterior supragnathal (pl . III, fig. 5 a- b) is of the cutting type wi th a fairly sharp Iow er edg e. It is in the form of a pl ate, 43 mm long , mesially divided by a vertical ridge into the hind portion 27 mm in length and the front portion 10 mm in length. The basal width of this r idge-lik e thickening is 6 mm, being 4 mm at top. It extends above the upper border of the elemen t an d forms a rod-like upper process gently curved inward. The .p oster ior supragnathal attains a maximum height of 25 mm within the ridge-like thick ening. Laterally, the anterior th ickened p or tion , 20 mm in l ength, forms a protrusi on limited by the ver ti cal p osterior margin. This protrusion corresponds w ith the la teral 'pos ter ior supragnathal process occu rr ing in genus Coccosteus and oth ers. Above this process, on the upper border on the element, is an impression to fit th e ligaments or muscles. The low er margin, seen from the side, .has the ou tline of a flattened letter "W", sh owin g two sharp bends. In agreement with this arr angement; the surface touch in g the infragn a­ tha] and occu rr ing on the mesial side of this element is narrow .and UPPER DEVONIAN FISHES 301

arched posteriorly, while anteriorly it widens out overlapping th e vertical ridge-like thickening and then ab rup tly n arrow ing again . Occurrence. - The Frasnian, horizon II in the Wietrznia Hill.

Genus Dinichthys Newberry, 1885 Remarks. - Genus Dinichthys was erected by Newberry to include species D. herzeri described by him. Dater , a n um ber of sp ecies was includ ed into the genus, 'Some of which su bsequen tly w ere referred to other genera, while many of the remaining forms are known only on very inadequate fragmen ts. In 1946, Dunkle and Bungart differentiated two groups of forms in the dinichthyids. On e of them, in those authors' opinion comprising D. herzeri Newberryand genus Gorgonichthys, is characterized by jaws with a Iblunt crushing edge, while th e other group, typ ically represented by D. terrelli Newberry and D. interm edius Newberry , displays sharp, cutting edges. These suggesti ons led Lehman (1956) to separate D. herzeri from the remaining forms, retainiog for it: the gen eric name of Dinich­ thys and creating the new genus to include the forms of the '"ter relli" group. Even such forms whose referability to gen us Dinichthys had not been questione d un til a shor t time ago, represen t various evolutionary stages and lin es of th is stock. To give a true !pictJure of the relationship of these forms it will, doubtl ess, be necessary to separa te a number of subgenera, or maybe, even genera.At tJhe present mom en t, however, Lehman 's other wise reasonable standpoint creates many p ractical difficulties. It eith er leads t'oa m echanical a ssignment of all forms, D. herzeri excepted. to the new genus Dunkleosteu s, engen dering th e p ossibihty of bh eir being subsequently shi fted from one to an otlher systematic unit, or left un­ classified. Both these possibilities seem obviously u nrecommendable to the p resent author Who is prompted tentatively to re tain th e ol d systematic arramgernen t of Dinichthys al terable in the future, after a d etailed revision 'based on st ud y of n ew materi-al concerning t'he now in ad equately known forms.

Dinich thys pu .> tu !osus Eas bmam, 1897 (text-fig. 8, 9; pl. IV, V; pl. VI, fig. 1-3) 1897. Din ich thy s pustulosus Eastman; C. R. E astmen, On th e relations ..., p . 38. 1898. Dinichthys pustulosus Eastman; C. R. Eastman, Some new p oints..., p. 748-­ 754, fig. 1, 2. 1907a . D ini chthy s pustulosus Eastman; C. R. Eastman, D ev on ic fishes ..., p. 130-13Q. pl. 12. 1918. Dinichthy s p ustulosus E astman ; L. H ussakoff & W. L . Bryan t, Catalog of fos­ sil fishes..., p . 50-53, pl. 12; pl. 13, fig. 1, 2, 4. J!l5fi. Dinichthy s cf. pustulosus Kulczyck i; J. Kulczycki, On th e parasphenoid ..:. p . 105----106, pl. 1, f ig. G ; p l. 2, f ig . 3. 302 JULIAN KULCZYCKl

Material. - Material consisting of two large cranial fragments, one [rom the Lower, the other from th e Middle Frasnian in the Wietrznia hill; fragments of the median dorsal, nuchal, p aranuchal and anterior lateral plates from the Middle Frasnian of Wietrznia, while an anterior sup ragnathal has been recovered from the lower Ch eiloceras beds, and two median dorsal fragments fr om the Frasnian ho ri zon III (Mantico­ ceras beds) in Kadzielnia. Description. - Head shield fragment from the Lower Frasnian of th e Wietrznia hill (fig. 8B ; pI. IV) comprises a complete nuchal and right pa­ ranuchal, part of left paranuchal, both centrals without anterior processes, medial part of the marginal and a small portion of the postorbital. This specimen has not been distorted and exhibits th e natural doming of this portion of the head shield. On the posterior margin, limited by th e nuchal, th e head shield is with a roof-like bend of about 130°. Farther to the front the doming grows gentle, more arcuate an d, beginning from the anterior suture of the nuchal hone, a shallow , ex tensive d epression is indicat ed . stretching to the sides of the supraorbital lines on the centrals. The nuchal is with a mark edl y regular outline which is almost perfectly symmetr ica l. Its free outer surface is of trap ezoidal form, 110 mm in height and w ith the basal width equal to about 180 mm, whe n m easured along the chord line between the m os t outlying lateral poin ts. The posterior margin, limiting the nuchal gap, is sligh tly concave. arched, provided with a centra'l ip rocess , 40 mm broad at the base. The an terior margi n, 68 mrn long, Iimiting bo th centrals, runs in a straight line across th e head shield, interrupted by only three sm all, sharp p ro­ cesses. The m iddle one of these wedges in between the two cen trals. while the two others are placed on th e side en ds of th e anterior nuchal margin '.'lith their apexes directed to th e middle of the r espective centrals. The lateral nuchal margins frontally uniting with the posterior central processes and to the hind with th e paranuchals, show a gently arcuate cou rse, on ly slightly brok en up in the area of contact with th e su ture between the cen tral and the paranuchal, In the hindmost section only , th e suture between the n uchal and paranucha! for ms two larger inden tations. Ventrally, th e nuchal is considerably narrower. Its posterior margin is 120 mm long. To the front the nuchal con trac ts centrally to 45 mm broad ening out again to 55 mrn in th e region of the anterior paranuchal margins. From this point the nuchal m argins extend in converging rectilinear lines directed to the front, producing an angle of 70°. and wedging in between both centrals. In the centre of this area an elevation rises from which stretches a faintly indicated. ridge laterally limited by likewise slight grooves. At a distance of approx. 80 mm UPPER DEVONIAN FISHES 303

from the anterior end there occu rs on the ventr al surface of the n uchal plate a depression (the "double sockets" of H eint z), sharply limited at the front as w en as at the back , 40 mm broad and 15 mm long and high. The bi-partition of the depression is in d icated by a slight n otch on the front wall togeth er with. a small m edian crest, and two, similarly small :'crests on the back wall. The posterior limit of the depression consists of a transversal ridge ("tr ansversal commissure" of Heintz), considerably more massive than in Dini chthys interm edius with a thickness of 20 mm. From its back wall, a small m edian crest p robably ending in a larger process , as w ell as tw o lower la teralcrests ex ten d on the 'cen tral proc ess of the p osterior nuchal m argin. At the transversel commissur e the nuchal thickness attains 30 mm. As m ay b e seen in the secti on ob ta ined by the breaking off of the ar ea oc cupied by th e left joint socket, the nuchal wedges in this area in to the deep pocket produced by th e m es ial para­ nuchal margin and forms the major bulk of the ridge (the "hind thickening" of Heintz) Which mesially Iimits the p osterior lateradcranial cavities. The paranuchal has the ou tline of a triangl e with truncated corners. The hind corner involves the joint socket. The shape of the latter ele­ men t suggests that the 'condyle m ust h ave b een th ick and r elatively short, The w id th of the joint sur fa ce m easured p ar allel to the axis of the joint is 22 mm, while the diameter of the curvature is 16 mm on the lateral margin and 14 mm on the m esial. It should be stressed that the axis of the joint is n ot placed hori zontally , as figured by Heintz in the r e­ construction of the skull of Dini chthys intermedi us (1932, fig. 35), but runs obliquely fr om ou ts ide of the top and front mesially downward and backward at an ang le of approx. 15° to the horizontal line drawn across the head shield. The lower lip of th e jo in t socket unites w ith the base of the joint process (lacking in the specim en under consideration): At th e base it is 14 mm wide and 6 mm thi ck. Quite close to the boundary between the outer m argin of the jo int surface and the mesial margin of the j oint process, the paranuchal margin prod uc es a m oderately conspicuous, sharp process traversed by the groove of the lateral line passing on to the anterior do rsolateral. At a small distance from this place, the posterior latera l margin of the paranuchal limiting the d eft between the h ead shield and body armor b en eath the jo int, forms a small list wh ich, with head lowered , is over'lapped vby iJhe front m arg in of th e pectoral gird le . Mesially, the upper Up of the joint surface form s a shamp , beak-like process overhanging the jo int cavity and at the same time closing UJP the pocket-like depression of the median paranuchal margin occupied by the l ater al part of the nuchal- plate. From this point, along the lateral part of the posterior nuchal margin, extends the fairly long shelf-like list, taking part in the limitation of tJhe anterior head shield 304 JULIAN KULCZYCKI ma rgin . This process resembles a simila r structure dn Pl ourtiosteus traut­ scholdi. The anterior paranuchal m argin con tacting with !!he central and marginal plates is irregularly arcua te . The outer p aranuchal surface is traversed by the sensory lime can al , by Gross 'referred to as tempor al, while He intz calls it margin al, and by an "occipital canal" (called so in th e nomenclature of Gross, by Stensio called the "poste rior pit-line") no t

Fi g. 8. - D inichth y s pustulosus, pa rt reconstr uction of head shield (dotted areas): A from the Middle Frasnian, B from th e Lo wer Frasnian of Wietrznia ; C central. N nuchal , M marginal, PN paran uch al , PrO preorbital , pta postorbital. UPPER DEVONIAN FISHES 305 reaching the suture with the central , In th e angle produced at t'he contact of these canals is the opening of the endolyrnp hatic canal. At a distance of 10 mm from the 8IPex of this angle, the te mpor al canal branches off passing onto the anterior dorsol ateral. The ventral surface of the nuchal in Dinichthys pustulosus is distinguished by the consider­ able width of the mesial step-like elevation forming a pocket for the later al nu chal margin and, together wi th it, p rod ucing the hind thick­ en ing of Heintz. This th ickening as compared to that in D. interm edius, is very m assi ve and characterized by considerable wid th, in creasing anter iorly (4~ mm above the joint socke t, 57 mm near the transversal th ick en ing, limiting the nuchal d epressi on) . Within the cen tral portion of the h ind thickening, along its lateral side, a broad but shallow depression extends ventrally, which, to the hind passes into a similarly wide and low blind canal, rapidly growing n arrower. Within the lateral wall of the hind head shield thickening, formed by the vertical paran uchal shelf, runs the endolymphatic canal. Its inner opening is placed at the an terior end of the above m entioned thi ckening opposite to the point of convergence of the nuchal, paranuchal and central sutures. The centrals show an intricate outline. Medially they are united by som ewhat flexuous suture. In th e h ind of the mesial part they are trans­ versally truncated by the nuchal element stretch in g far to the front. The posterior tongue-shaped processes of the centrals adjoin the fore part of the lateral nuchal margin, Laterally to these processes wedged in between the nu ch al and Paranuchal, the m argin of th e central after a slight bend resumes its rectilinear course sidewar d to touch the genal plate where it curves and then m esially runs on arcuately forward. The width of the centr al is 92 mm a'S measured fr om the poin t wh ere the two centr als and nuchal come together to the point of converg ence between the cen tral, genal and paranuchal. The length is 97 mm, as measured from the hin dmost point of the posterior process laterally to the frontal m argin of the an terior p rocesses which are missing on the here consid ered specimen. On the external surface of the cen tral pl ate, supraorbital and central , se nsory line canals are visible, converging to the m iddle of the element. Th e latter of th ese lin es projects somewhat fa r ther across th e [p late inward in relation to the supraorbital line. Besides these two canals, an isola te d section of the sensory line (the "Restk an al" of Gross ,the "central pi t-line" of Stensio) is also fou n d on the central. After running inwards in a tr ansversal xldrection to the mid dle of tlhe element, this line bends backwards without, however, attain in g the margin ad jacent to the nuchal. A considerable portion of th e ventral surface of the central is overlapped by the plates encircling it so that its free surface is confined to the middle area. In the studied

Acta Palaeontologlca Polonlca - Vol. IIj4 306 JULIA N K ULC ZYCKI specimen this is 71 mrn in wid th and 74 rnm in length. This area is occupied by a shallow, bow l- lik e d epression, withou t sh arp lateral limits, elsewhere surrounded b y a semicircular thi ckening. At r the boundary w ith the nu chal plate there is a groove. It forms an embayment near th e end of the paranuchalat the opening of the endolyrnphatic canal. Far ther th is embayment opens into the posterior lateral cavity and extends avera short distance al ong th e suture betw een the central and the .paran uchal. The only preserved pa rt of the marginal is that oc cupying a mesial position in r elation to the lateral line groove r unning across this element. Along the anterior par-t of this Iin e, on the ven tr al surface of the plate, s tretches a ridge-like thickening which posteriorly low ers down and becomes sharp-e dged. It cons titutes the hind part of the "la te ral consolidated par t" of H eintz. At the boundary betw een the marginal and the postorbital li t is 30 m m long . A large area of the external surface is smoothed ou.t, undam aged tuber cu la tion is to be found on the centrals orrly. The tubercles are very minute, be low 0.3 mm and densely disposed (100 to 200 over 1 em"). The parasphenoid, alread y d escribed in d etails b y the p r esent au thor (Kulczyeki, 1956) w as found together with the ab ove con sidered head shield fragment. From the same element in Dini chthys te rrelli it di ff ers primarily in t he pr esence on the dorsal surfa ce of a small tube r culated area. The other differences main ly concern proportions. The other cranial fragmen t (fig. SA; pI. VI , fig. 2) from the Mi ddle F rasnian beds of the Wi etrznia hill (horizon II), belongs to an individual w it h similar, perhaps so mew hat lesser dimensions, and is characterized by less regular shape of bones . This is particu larly so in the case of sutu re b etw een the two centrals which displays a marke dly flexuous cour se, also in that b etwe en the centrals and the n uc hal which is eq ually zigzag and less sym metrical than in the above described spe cimen. The proper­ tions of the respective b on es a re also so mewhat different. For example, the n uch al is h ere 'relatively short and wide, b eing 90 mm long as measured dorsally adong the med ian li ne, w h ile the w id th of the anterior margin is 75 rom, the length of the posterior m argin being p r ob ably about 180 rnm. The rather b ad state of preservation of the stud ied specimen m ak es impossib le a closer comparison of proportions in the pa r ticu lar pla te s, it would seem , however, that the shor tness of the nuchal is its chief characteristic. The ventral leng th of the nuchal is 100 mm as ag ains t 130 mm d isplayed b y the lower Frasnian specimen, The d istance from the base of the joint process to the anterior n uchal extremity is here 155 rnm, b eing 175 mm in the Lower Frasnian specimen.Corespond­ ~ngl y , the distance from th e point of divergence of the senso ry line on

I , UPPER DEVONIAN F ISH ES 307

.th e marginal from the m edian h ead line, measured .on the cranial . cu rvature, is 140 and 130 mm respectively, The proportions oftJhe cen trals were markedly similar or even the same as obs erved in the .individua l fr om th e Lower Frasnian. The w idth and length (w ithou t th e anterior process) a r e, 1n botJh leases, 70 mm. The length of the anterior 'process is not less than 30 mm. The considered specimen is with an almost p erfectly pr eserved marginal. It is a relatively narrow bon e, w ith wid th n ever exc eeding 50 m m and length up to 110 mrn, forming the lateral m argin of the craruia l dome with the poster ior later al cormer. The latter p rojects far to th e side (the d istance from its apex to th e median h ead line, m easured on the cu rvature, belingajplprox . 200 mrn) and does no t d isplay the presence of an independen t rp ostmarginal element. The lateral border of the marginal is in curved to produce an angle at the m iddle b one level. Its shape indica tes that, similarly as in Coccosteus and other species of Dinichthys, the union of tJhe crandal d ome wdth the bones in the cheek area w as not intimate. The boundary with the postorbital is an teriorl y i nd en ted where the sensory lin e groove p asses from one bone to the 'next. This is similar to what h as been observed lin the before d esordbed specimen and in the fragment of a North American h ead shield (Eastm an , 1898, fig . 2). On the lower sur face, from the apex of the process formin g the exterior angle of the skull to the middle of the element and then parallel to the sensory line, stretches a thickening. It tis low at dts beginning , but gains in height as it advances; sharp-edged in its h ind part, and [particularly so in bhe central rr egion. This lis the poster ior p art of the so called lateral consolidated part. It displays a frontal heigh! of 25 mm. This thickening passes on t'o the postorbital 'wher e, about the middle of the element it tapers into a rather narrow crest, inwardly inclined . On tJhe ventral slide it exhibits a concavity (the double sockets) posteriorly steeply limited. The middle part of the pos torbital be ing damaged, it is im possib le to ascertain whether the median cranial process existed at all and how it was shaped. Hereabout, along the course of the 'cen tral groove on the external surface, the trdangular, shelf-like \process of the central plate is wedged into the inner border of the mar gimal. On the ventral surface, th e suture between the marginal and the postorbital has a similar course as that in Dinichthys inte rm edius. The course of th e sensory 'line grooves displays some individual var iations as compared with the Lower Frasnian head shield . To say , th e cootral groove does not on the inward side gobeyoll'd the line indicated by thecourse of the supraorbital groove. The so called vestigial groove or the median 'Pit-line lies very close to 't!he central groove, cu r- 308 J ULIAN KULCZYCKI

ving backward it reaches the margrn of the central plate contacting with th e nuchal. In these details the here d escrib ed specimen closely resembles the head shield described from North America (fig. 9). In addition to these major samples, two smaller cranial fragments have been found dn the Middle Frasn ian of the Wietrznia hill. An incomplete nuchal represents an individual at 'least three and a half times larger than that from the Lower Frasnian. On tJhe ventral side it exhibits a con cavity (the double sockets) posteriorly steeply limi­ ted . On 1lhe p osterior margin there .is a surface for the process of the paranuchal reachmg far inwar ds . A deep notch in 1lhe frontal part of the lateral margin indicates that th e paranuchal overlapping the nuchal, particularly so on tJhe dorsal side, w as itself wedged into th e margin of that element by means of a stout shelf . A fragment of the joint socket area of the paranuchal is, on the contrary, to be referred to a considerably' smaller individual. The width of the joint socket, measured parallel to 1:Jhe axis of the joint, is scarcely 13 rnm, while the diameter of the curvature is 9 to 10 mm. In spite of the sma-ller dimensions of that specimen the size of the ornaanenting tubercles and its density do not diff er from those common in larger individuals, which indicates that the size of tubercles d id not al ter with the age of the individual. Of the jaw apparatus belonging to the here studied form our m aterial only contains the r ight ran ter-ior supragnathal collected from the lowe r Cheiloceras beds of Kadzielnia (p], VI, fig 1). Its shape is that typical of gen us Dinichthys. We ma y distinguish here an anterior part which is narrower, being only about 15 mm wide, and a wider posterior (lateral) one, abou t 30 mm in width . Both these parts m eet at an angle of nearly 90° forming a ridge, i. n the lower portion made conspicu ous as a rounded crest. In the upper ,pa,rt it iis replaced by a triangular nearly plane sur face. Along this r idge th e anter ior supragn athal is thickened and elongating do wnwards forms a tooth-like process, triangular in section. Another tooth-like process, though much smaller, occurs at the lower end of the posterior ed ge of the el ement. The upper end ad' th e anterior par t continues as a lob e-like process, curved inward. The presence of a series of "denticles" on the vertical edge of the tooth-like process is an ind ex rch aracter permitting the assignmen t of th is jaw element to the remains of Dinichthy s pustulosus. The height of the preserved p ar t of the anterior supragnathal is 42 mm. Another 10 mm must be added to this figure to make up the missing apical pick of the tooth-like process. Of the bod y armour the Holy Cross . M ts, m aterial contains fOUT fragments of the median dorsal and on e orf the anterior lateral. The .most complete specimen of .the median dorsal (pl . V) has been UPPER DEVONIAN FISHES 309

/ "

. ~. / f ...... f \ PN N ..../ i \, .,._..././..•.., . \ ;" ~ \ r / ../,,'" -c. ./ .....'\ . ./ ......

/ ' ;"...... "

N \ " ) i ". \ \,

<...... ,...... ,..../,..._,.\'-" ...... ~ _ • _

Fig. 9. - D ini cht hy s pustu!osus; sketch drawings of fragmentary North Am erica n h ea d shields , after E astman. 1898, fig. 1; 1907, pI. 12. L egend - as in fig. 8. JULIAN KULC ZYCKI

recovered from the Lower Frasnian of the Wietrznia hill, together with the descr ibed above fragment of skull. It is stro ngly curved at an angle of 110°, without, however, forming an edge. Ven trally th ere is a strongly developed keel, attaining in th e hind p art a h eight of 60 mm and a length of over 180 mm at the base. The p ro cess of - the k eel, not preserved in th e studied spe cimen, w as placed lin a more vertical p osi tion than th at in Din ichthys intermedius, meeting with the external surface of the ele me nt at an ang le of 110°. From its base, clearly indicated ridge-like thickenings extend to the side for ming an angle of nearly 90°. They go on arching laterally and growing flatter and broad ening ou t to disappear completely at a distanc e of abo ut 80 mm fr om the keel. The dorsal surface is t uberculated fin ely but apparently no t so densely as in the above d escribed fragment of th e head shield . Very faint traces are discernible of the sensory line grooves (pl. V, fig. 2), running obliquely on the dorsal surface of the median dorsal from the sid e and front, inwards, to the hind [p art of th e surface, wi thou t, however, attaining it. The other median dorsal fra gme nt collected from th e Middle Frasnian of the Wietrznia hill is in a still more fragmentary state of preservation, being brok en UiP into minute f-ragments. It belonged to a far larger individual and may 'have been less strongly transversally curved. On the outer surface the tubercles show a m ost uniform and very dense ar ran ge ment. The two remaining median dorsal fragments, recovered from the Upper Frasnian (Manticoceras beds) in Kadzdelmia, are still more in­ complete. They are referabl e to Dinichthys pustulosus on the character of ornamentation. The an terior later al fragmen t (pl. VI, fig . 3) corresp onds to that par t of the anterior margin which was bent at th e level of the posterior • lateral corner of th e head shield. Similarly as an D. intermedius, above the bent 0Il1 the front m argin of th e anterior lateral, there is an elongated excavatio n, more closely limited anteriorl y, overlapping the hind margin of the head shield, when the h ead is lowered. Farther downward, the fron t margin of the antenior lateral widens out forming the base of the median p rocess connected with the in t'erolateral. Here a characteristic knob is to be noted. On the outer surface the tubercles are arranged with relative scarcity (less than 100 over 1 em"), similarly as they are on the median dorsal from the Lower Frasnian. Remarks. - The shape of the anterior lateral, th e median dorsal with its characteristic keel, th e anteri or supragnathal, as well as the general arrangement of bones in the head sh ield , and the free cheek Mea, all indicate that the above described fossils are doubtlessly referable to genus Dinichthys (as interpreted above). Similar ornamentation is UPPER DEVONIAN FISHES d isplayed besides D. pustulosus Eas tman, also by other representatives o f this genus such as D. oviformis Gross from the Eifelian of the Rhine provin ce, and D. magnificu s Hussak of & Bryant from Genesee ;in the state of New Yo rlk. Dinichthys ov iformis ds known on det ach ed fragmen ts of the Ipar a­ nuch al, anterior dorsolateral and median dorsal described by Gross (l933b, 1937). The two first mentioned spe cimens differ from the remain­ ing on es as well as fr om the Hol y Cross Mts. sp ecimens in larger size an d less dense arrangemen t of t ubercles on the p eriphery. One specimen of the median dorsal of D. oviform is is with rounded , very clos ely disposed tub ercles, and is not provided wi th thickened r idges r eaching to the base of the crest. · This specimen is doubtlessly referable to a different form th an that described above. The other m edian dorsal specimen of D. ovifonnis in charact er of or nam en tation , preserved ov er a small area only, resembles more closely the Holy Cross Mts. specimens. but differs from them ,in the taperin g hind end and less strong trans­ 0 versal curvature forming here an angle of 120 • Dinichthys magnificus differs conspicuously fr om our form both in th e shape of the whole h ead which narrows strongly anteriorly and wid ens out poste.riorly, and in the outline of the various b ones. The resemblance of the Holy Cross Mts. speoimens with D. pustu­ losus is, on the oth er hand, obviously striloing. Besides such characteristic features as the presence of a number of denticles on the anterior supragnathal, of vestigial sensory lines on the median dorsal, the outlines of the various b ones of the h ead shields in both - the American and Holy Cross Mts. specimens - are identical so much so as to dnclude such a detail as for example the bent of the su ture between the central and t'he marginal, situated to the si de of the hind tongue-shaped process li n the central pa rt. Som e existing unimpor­ tan t differences do not appear to th e present author as exceeding the range of individual variation which, on evidence of specimens fr om Wietrznia and more particularly so of the North American head shields 1S rather strong within this form and is mainly associated w ith irregula­ rities of growth and ossification. The only differences, possibly obs erv­ able, between specime ns from North Am erica and those from the Holy Cross Mts. are as follows: a) somewhat coarser tuberculation in the Arne­ r ican specimens; b) the presence of an additional branching of the sensor y lin e groove on the central plate (the "Restik anaa. " of Gross, or the "median pit-line 0' of Stensio); finally possibly also less clearly marked sensory line grooves on the median dorsals of the Polish sp ecimens. This last difference, however, has not been quite doubtlessly ascertained by the writer since no adequate illustrations of the American median d orsal 312 JULIAN KULCZYCKI specimens were available besides a schematic drawing. Inasmuch, how­ ever , as these dissimilarities are of no great importance and pertain to characters displaying strong variability, it is probable that we have here a geographical variety of D. pustulosus representing a somewhat more advanced evolutionary stage. D. pustulosue was established by Eastman (1897) on evidence of a fragmentary ventral plate, probably the anterior ventrolateral , and of a somewhat more co mplete anteri or d orsolateral , both distingu ished by a "finely tuberculated style of ornament" .As fi gured dn the orig inal description , these tubercles, however, are large and scar ce, ra ther approaching the ornamentation of D. tuberculatus. They do not fit into the description of other specimen s of D. pustulosu s p ublish ed subseq uently either by Eastman himself or Ib y other authors. This is most likely 8. result of t he lack of precision in fig uring them , since E-astm an writes that: "... the ar tist has represented these somewhat diagrammatically..." . Hence, during comparative studies of material the p resent writer relies on fuller descriptions of more complete matenial later published by Eastman (1898) and by Hussakof and Bryant (1918) rather than on East­ man's type. The presence of a branch of the sensory line g roove on the central plate, as mentioned here above, is th us not regarded by the present writer to be a constan t feature of the American specimens; in any cas e its b ehaviour varies consider ably. On evidence of the above da ta, D. pustulosus corresponds to a p ri­ mitive form of moderate dimensions , distinguished a) by markedly fin e tuberculati on which is absent from very narrow s tr ips only al ong the margins of bon es; b) b y strongly domed h ead shield; c) by somewhat flexuous cours e of sutures ; d) by a r elatively long con tact of th e centrals along the m edian line; e) by an anteriorly bluntly truncated , fadrly long. trapezo id al nuchal ; f ) by the presence of the me dian pit-line (the "Restk anal" of Gross ) on th e cent rals, and of traces of the sensory lin e grooves on the m edian dorsal; g) finally b y th e !presen ce of vestigial tooth denticles on the Infragnathal and anterior supragnat!hal. In his p aper on the Brachy thor aci f rom T af ilalet in Marocco published in 1956, Lehman r ef ers to the view h eld by 0rvig , acc or-ddng to whi ch D . pustu losus is no t believed to be a dinichthyid. Unfortunate ly, Lehman does not state any evide nce for what he postulates besides mentioning the fossil r emains describ ed by Hussakof in 1942. In what t hese are concerned, the views advanced by the a uthors mentioned ab ove seem reasonably admissible. Neverthel ess, as _ suggested b y 'ch aracter of ornamentation, neither do Hussakof's specimens described in 1942 belong to D. pustulosus. In what r egards other American materials known to UPPER DEVONIAN FISHES 31 13 the present writer from literature, and still more .so in regard to sp eci­ mens from the Holy Cross Mts., it is notexcluded tJhat a fuller knowledge of the here considered form in particular and that of Djnichthyidae in general will necessitate the subgeneric or even generic separation of D. pu stulosus, hut will also confirm beyond doubt dts association with 1he Dinichthyidae. Occurrence. - D. pustulosus has thus far been recorded from the Middle Devonian (Hamilton) of the states of Wisconsin, Iowa and Illi­ nois, also from the UPIPer Devonian (Genesee) of Indiana, Kentucky and New York; now the range of this form must be extended throughou t the Frasrrian and the lower Cheiloceras Famennian horizons in the Holy Cross Mts.

Dinichthys denisoni ;; n, Sip. (text-fig. 10; pI. VI, fig. 4; pl. VII) Material. _ . Three specimens of t he median dorsal, two of which come from the Clymenia beds oil' Galezice and the third one from the low er Cheiloceras beds of Kadzielnaa in Kielce. ' Diagnosis. - A rather small representative of ge nus Dinichthys with bones unornamented, with tJhe median d orsal characterized 'by a straight course of the lower margin of the keel, like it is i n Dinichthys? jefferso­ nensis Branson & Mehl, but differring from it fn gr eater height of keel an d in presence on apex of the carinal process of a typical spoon-like excavation. Holotype. - A n early perfect median dorsal (1. G.) (pl. VII, fig. 2 a-c). Descri ption . - All the specimens of the me dian dorsal plates are of p retty the same size, Jndicating that they represent the average di mensions of 'an adult individual of this form. The best preserved spec im en (pl . VII, fig . 2 a-c) from the Clymenia beds, is 135 mm in width, when measured along the chord line, a nd about 150 mrn, when mesured along the curvature. It is gently domed and relatively strongly elongated, attaining in the median line a lengt h of above 105 mm. The hind, gently rounded margin displays centrally, above the base of the carin al process, a rather small, semicircular process usually missing in representatives of genus Dinichthys. The !keel, as m en tioned in the above diagnosis, Ii's characterized b y a r ectilin ear margin stretching to the apex of the caoinal process and attaining a length of 125 m m. 'I'he vcarinal pro cess is slender and fa irly long (60 mm in length and 15 mm in width) It starts wi th a slightl y broadened ou t base without any branching

5 This species is dedicated to Dr. R. H. Denison, Curator of the Fossil Fi shes Depa r tment in the Natural History Museum of Chicago. :314 JULIAN K U LCZYCKI

thickened ridges. Seen in profile its outline is gently arched. In th e upper :portion of th e posterior surface a rather small ri b runs along the med ian line extendin g to half the length of the proc ess. A thi r d, low er part of th e process is occupied by a characteristic sp oon-lik e excavation. The hind m argin of the carinal p ro cess m ee ts with th e dorsal surface of the eleme n t at an angle of abou t 120°. The other specimen (pl. VII, fig . 1 a-c) , collected at the same site. di ffers from that d escribed above in somew hat larger d im ensions. Th e carinal process here attains nearly 80 mm of len gth. In relation to the dorsal surface of the element it is placed at em angle of 125". The sp ecimen (pl. VI , fig. 4) from lower Cheiioceras shal es of Kadzielnia is strongly cern­ pressed. Its state of preserva­ tion, h owev er, permits to ascer ­ tain that the shape of th e crest Fig. 10. - Dinichthys denisoni n. sp., and of the carinal pro::ess were longitudinal section of the median dorsal w ith d iscernible outline of crest. Outline the same as those in specimen s of less complete specime n sh own by fro m the Clym en ia beds of Ga­ broken line. lezice. Si milarly, the vaulting of the bones is gentle. The carinal process is here 75 mm long. Remarks. - In r egard to the shape of the crest, the here considered for m r esembles D. jeffersonen sis, but differs from it in greater heigh t of crest and a more characteristic sh ape of the carin al precess. Occurrence. - The Famennian, lower Cheiloceras beds of the Kadziel­ ni a hill in Ki elce, and th e Clymenia beds of G alezice.

Dinichthys ceterus" n . Sip . (pI. VIII, fi g. 1 a-b ) Material. - One specimen of the median dorsal. Diagnosis. - A small form characterized b y lack of ornamen ta tion an d the nearly h orizontal position of the carinal p ro cess. Holotype. - A fragmen t of the median dorsal (lVI . Z .) (pI. VIII. fig. 1 a- b). Descrip tion. - The preserved fragment of the median dorsal 120 mm in len gth, comprises the median part of the elemen t including the crest. The crest has a m aximum height of 30 mrn, while its length is over 150 mm. The carinal process is slightly curved from the pl an e, so that it projected very conspicuously beyond the hind margin of the median dorsal. A transverse thickening a t the base of the carin al process is no t

• The specific name is derived from the Latin w or d cet eru s - other. UPPER DEVONIAN F ISHES 315

indicated on the dorsal surface of the element, thus differing from the 'corresponding thickening in D. pustulosus. From the latter form, as w ell as from D. denisoni and other representatives of the same genus D. ce­ terus differs also in the presence of a sh arp hind margin of the thickening at the base of the carinal process. . Occurrence. - The Famennian, Cheiloceras beds of the KadzieLruia hill in Kielce.

Dinichthys d. tuberculatus New/berry (pl. VIII, fig. 2, 3) Material. - Several,chruracteristically ornamented fr-agments from the Clymenia beds of the F amennian in Galezice, From this material fragments of the preorbital, margimal and para­ .nuchal have been recognized. They are, however, 'so Iragmentary that they give no clue to the mere signifi cant structural details. A more complete fragmen t constitutes the supposed hind portion of the right posterior ventrolateral. In addition to m inute, t ypically ornamen ted head sh ield fragments, an impression h as b een found of the r ight infr agnathal mesial surface comprising a small [piece of bone corr esponding to the anterior lower portion of that elem ent, Description. - The infragnath al (p l. VII, fig. 2) w hose b roken off , probably posterior end left no tr aces on the rook surface, was foun d to be over 200 mm long. Its h ind portion constituting the "blade" fus ed wi th the upper part of the arched jaw, is of a rather robust appearance. It is 65 mm wide at the posterior end and 38 mm forward where the "blad e" passes into the fore ;part. The latter, 109 rom long, occupies ab out one half of the entire length of the elem en t. Along the lower margin there is a thickening posteriorly passing, without sharp limita­ tion, into the back, blade-like part and projecting forward along the anterior m argin and to the top passing in to a larg e toothed process. As compared with D. intermedius and other r epresentatives of this genus, the h ere described infragnathal is, among ot hers, disti ng uished by th e rather sm all height of the above m entioned thickening. It is her e but 15 m m high against 45 mm of entire height shown by the anterior part of the elem en t. The remaining portion of the for e part of the infragnathal, lying above the thicken ing, ·is gently ar ch ed and smooth . In h alf len gth only, on the functional edge, there is a thickened ridge for ming the second, smaller toothed [process. It is nearly 20 rom high and 5 rom broad, and its position is very characteristic being equidista nt from the anterior and posterior ends of the functional part. In this point it differs from other species of Dinichthys. Its upper cutting edge, contrary to what we see in D. intermedius, D. terrelli 'and others, in ,3 16 JULIAN K ULCZYCKI

,D. newberryi Clarke, particularly so.d isp lays no incision between the two toothed processes but runs almost rectilinearly to the back, where it descends abruptly near the boundary witlhthe posterior part of the infragnathal. The upper cutting edge rises slightly in the region corresponding to ' the smaller toothed process only. As a whole the infragnathal, when seen from above, is gently sigmoidal, somewhat more $0 at the posterior, inwardly directed end. The supposed posterior ventrolateral is 101 mm broad. Its outer su r face shows an ornamentation consisting of irregularly disposed, moderately large tu ber cles, with diameter occasionally up to 1.5 mm, mostly rounded, domed, frequently quite flat, more seldom of a conically pointed shape. Remarks. - The shape of t h e

Dinichthys? sp. (pl. VI, fig. 5) An anterior dorsolateral, probably referable to Dinichthys, has been recovered from the lower Cheiloceras beds of Kadzielnia. It must have belonged to an individual of moderately large size. The unornamented outer surface is traversed by two grooves of the lateral line. One of them projects from the anterior margin of the bone where the sensory line passes onto the paranuchal above the joint, the other start'S at a distance of about 40 mm from the anterior margin of the bone and is directed somewhat obliquely upward and backward. As stated by Heintz (1932) .similar secondary grooves are also observable in D. intermedius, where 318 JULIAN KULCZYCKI

as many as three of them occur occasionall y and are not in the least related to the line extending over the median dorsal in Coccosteus and oth er similar forms. According to data published by Eastman, there is also an additional groove on the anterior dorsolateral of Dinichthys pustulosus. Its course, however, is different, resembling that common in the Holonemidae . In consideration of this character the ab ove men­ tion ed anterior dorsolateral cannot be regarded as an element of D. pustulosus from the same beds, w ith the ornamentation worn off. The free area in this sp eci men exceeded 10 mm in width. The length of the preserved fragment was 120 mm, indicating that the entire length must have been more than that. A portion of the anterior margin of the free are a situated above the condyle meets that p ortion lying below the condyle at an angle of 126°. The whole bone is gently domed, Its anterior margin is bent at an angle of 150°. The condyle must have been large and very robust. It is 33 mm wide at the base an d 17 mm thick at the mesi al border. To the rear and downward it gently passes into a thickening projecting, as in D. intermedius and probably in others, on the inner surface of bone. The thi ckness of the here considered sp ecimen attained 17 mm in the lower p art of the anterior margin, being 7 mm at the back. Remarks. - The specific p osition of this specimen cannot be now establishe d. It has been already mentioned abo ve that it is not an abrade d elemen t of D. pustulosus. The relative robustness of the anterior dorsola te ral structure as compared with the slightly more delicate struc ture of the describ ed m edian dorsals sp eaks against its relationship with either of the two other forms recorded from the same beds.

Brachythoraci gen. et sp. indet. (~) (pl. X II, fig. 2) An an terior dorsolat eral , more closely in ::letermi nate, was found in the Clymenia beds of the Famennian in Galezice , which h ave also yielded remains of Ti tanich th ys, Pachyoste us and Dinic hthys. It is a fairly well prese rved plate wi th an almost complete condyl e. It is gently arched transversally an d has a narrow free area. Most pa rticularly na rrow was that pad lying beneath the sens ory line groove hori zontally intersecting the bone. The width of the free area is 45 mrn when the len gth of th e element is at least 85 mm w ithout the condyle which strongly protrudes bey ond the anterior m argin of th e bone. On th e ou ter surface of the down Slide, th ere is a large plan e overlapped by the an te rior la teral. To the front of the latter element, along the anterior margin of the anterior dorsolate r al a hi gh crest projects (17mm). QA,\theUIPPeT. margin of the specimen we can note a small portion UPPER DEVONIAN FISHES of the surface serving fair attachment with the median dorsal. Tho condyle is robust, 30 mm in length (measured parallel to the axis of joint) and 15 mm in thickness at the outer margin. Mesially, the condyle grows narrower tapering into a point. From the condyle, on the inner surface of the bone, a crest extends with hei ght first equal to the length of. the condyle and abruptly lowering to the rear, though s till discernible at th e posterior border of the bone as a slight elevation. The outer surface of the anterior dorsolateral ds perfectly smooth, without any ornamentation. The bent in the anterior m argin near the condyle, a's w ell as others characters speak against' their assignm en t to any of the mentioned genera.

Genus Titanichthys Newberry, 1885

Titanichthys kozlowskii 7 n. SiP. (text-fig. 11; pl. IX, fig. 6) MateriaL - Two head shield fragments, of which one comprises an incomplete nuchal and the leftcentral,vlhile th e oth er probably constitutes a part of the paranuchal. Diagnosis. - A relatively sm all form in which the centrals are large and extending far to the back and sides, thus resembling T. clarki Newberry and T. agassi zi Newberry, and differing from T. termieri Lehman. The scheme of the sensory lin e grooves on the centrals probably that' in the holotype of T. termieri. The nuchal broad at the front as in T. agassizi or in T. termieri. From T. agassizi it differs in another shape of the depression (double sockets) on the ventral surface of the nuchal, which is here characterized by larger posterior embayments, projecting far backward, while the anterior ones are smaller. Holotype. - A fragment of the h ead shield comprising a portion of the left central and nuchal plates. Description. - The nuchal had the outline of a low equilateral triangle with rather st rongly con cave base. Medially flat, its elongated posterior lateral portions are gently curved sideward and d ownward, so that as a whole this element W 3S slightly arcu ate . Very thin in the fore part, it attains a cons iderable thiokness (26 mm) in the h ind part, producing an arched thicke ne d ridge beyond which is the broad shelf of the posterior margin. Behind this thickening, in the central part, is an extensive double d epression (double sock ets) with a four-lobed out'line, m edially separated by a high crest. At the baok it is limited by a moderately high buckle-shaped crest. E.3Jch 0'£ the double sockets attains

1 This sp ecies is dedicated to Prof. R. Kozlowski, Head of the Instltuta of Paleozoology of the Polish Academy of Sciences. 320 JULIAN KULCZYCKI its maximum depth (approx. 20 mm) in the anterior media n par t and is divided by a thickened elevation of the side wall into two embayments. Of these the anterior projects farther to the side than the posterior one directed to th e rear. On the anterior border the excavation is up to 58 m m in width, its medial length is 40 mm arid it overlaps th e central process of the posterior nuchal margin. The process is over 15 rnm long and 40 mm broad, Ven trall y, beyond the posterior m argin of the double

I ~ I 1/ , 1/ \\ I II \\ I II . ~ I \\ I • I I I I I c I I l0'0~~K ------""" ' " " "" " <, N " -, -, " -, "" ,/ ------Fig. 11. - Tit an ichthys kozlowskii n. sp ., d lagramm a ­ ti ca l draw ing of poster ior part of head shield. Con­ tinuous lin e - li m its of preserved fragme nts; b rokan line - supposed course of suture be tween the nuch al (N) and the cen trals (C); d otted ar ea - p ar t 'of specimen w ith damaged surface.

ockets, the process is striped by sagittal r ibs, of which three - a cen tral cne and two laterals - are more conspicuo usly indicated as small crests. T he length of the preserved par t of the nuch al on the median line is 98 mm . The d istance between its poster-ior lateral ends projected far th est 10 the side exceeds 220 mm. No natural margin has been preserved on the central, but it must have been a very large bone reaching far to t he rear. The p re served fragment is 115 mm in length and 70 m m in width. Actually the central m ust have been a t least twi ce a's broad and considerably longer. Besid es th e much erased traces of the terminal parts of the supra­ orbital and cen tral lines we may note here a ch aracteristic independent branch of the sensory canals (the "Restkanal" of Gross, the "central p it" of Stensio), running from the middle of the .bone mesially an d UPPER DEVONIAN FISHES somewhat to the back , No occipital line is observable, s uch as present in Titanichthys agassizi, bu t its lack here may be due to the state of preservation. The b ad state of preservation does not permit a closer description of the supposed paranuchal, It is a thin 'bone, d isplay ing a more important thickening (32 rom) dn the mesial p art only , on w hich two impressions are to be seen ventrally, serving perhaps as the means of attach ment to the nuchal. This thickening probably cor responds to the "hind consolidated part" of H eintz. On the p osterior end of this th ickenin g there is an excavation w hi ch m ay be the joint sock et. It is rather large though shallow. Its length (measured parallel to the supposed axis of joint) is 50 m m, while the width is 30 rom. Remarks. - The characteristic features of the above d escr ib ed sp ecimens are the relative thinness and fineness of b ones common in the genera Titanichthys and Gorgonichth ys. Both genera are icharac terized by relatively large centrals . In Gorg onichth ys the shape of the joint socket apparently resembles the excavatio n presen t on the supposed p aranuchal. Unfortunately, w e lack closer d ata conceming this d etai ls in Titani::hthys. The assi gnment of specimens from the Holy Cross Mts. to the latter genus is suggested foremost by the sh ape of the ventral surface of th e nuchal w hi ch h as the characteristic form of a double excavation . The p resence of a ves tJigialcanal on the central, miss ing in a well known representative of the Gorgonichthys is a less diagn osti c featur e. As may be seen from the draw in g pu blished by Dunkle & Bun­ ga rt (1940), this region is differen tl y shaped in Gorgonichthys. D ifferen t proportions and details, for example the stronger d evelopment of the posterior embayments as compared with t'hose in T. agassizi, anoth er behaviou r of the centrals than those in T. termieri, also the absence of the occipital canal, finailly a different outline of the nuchal, as compared with T. clarki, - all indicate that we are here dealing with a new species of Titanich th ys. Occurrence. - Clym en ia beds of the Famennian in Galezice near Checiny,

Genus Stenosteus Dean , 190) Stenosteus? sp. (pl. X I, f ig. 4) The Cheiloceras (Lower F amennian) b eds of the Kadzielnia hill have yielded a rather small fragment of the infragnathal, referable to some rep r esen tative of Selenosteidae. It comprises the hind (por ti on of the fu nctional part together with an impression on the rock of the for e portion of the posterior blade. The functional p art of the infragnathal

Acta Palaeon t ol ogl ca Polont ca - Vol. n l4 21 322 JULIAN KULCZYCKI in this form was narrow and elongated. Its height was 8 mrn, While the length of the preserved fragment cons tituting no more than half of the en tire functional part was 21 mm. The upper edge of the functional part is provid ed with a single row of m inute knotty d enticles rather scarcely arranged (4 cu sps over 10 mm). A sim il ar single row of denticles also occurs in genus Rhinosteus, but tJh e deruticles there are more d ensely arranged and th e height of the entire functional part is considerably greate r. In connection with th e above feature s this poorly d efined specimen is by the p resent writer tentatively ref erred to genus S te ­ no steus.

Genus Pachyoste us Jaekel, 1903 Paziiuosteus bulla Jaekel, 1903 (fig. 12, 13; p I. IX, fig . 1-5) 1932b. Pachyosteus bulla Gross; W. Gross , Die Arthrodira Wildungens, p. 16-111, fig. 4 a-c, 5 a. 1933b . Pachyosteus bulla Gross ; W . Gross, Die Wirbaltiere..., p. 35.

Material. - Five specimens, one of them, comprising a large fragment of the head shield and shoulder girdle, contains a number of elements th us far unknown in th is genus, to say: the iPostsuborbital, postmarginal, an ter ior supragnathal, posterior supragnathal, probably the parasphenoid, the anterior lateral, anterior median ventral, interolateral, anterior ven trola teral. Other elements preserved almost complete in this specimen are the preorbital, suborbital, rostral, anterior part of the pineal and th e infragnathaI. Description: - The head shield as a whole is broad, strongly domed, 'blun tly termina ting at the fore end. The preorbitals , however, do no t form such sharply indicated anterior later al corners, as th ose reconstruc­ ted by Gross (1932 , fig. 4A ), in conseq uence of which the front of the head is more rounded . The rostral is rath er small, 14 mm long and 16 mm broad, its outline approaching that of an equilateral triangle with rounde d corners and gently incurved sides w hi ch m eet the preorbitals. The arc ua te an te rior border curves dow nw ard. The pi neal is approximately of the same width as tJhe rostral. The distance fro m the pmeal exc avation to the con tact with the rostral is 14 mm. The large orbited cavity, over 50 mm in d iameter, is limited by four bones. The preorbital, abo u t 50 m m long and 25 mm broad , do es n at differ in its outline from the corresponding element in Wildungen specimens . The postorbital is with damaged m esial and hind margins, UPPER DEVONI AN F IS HES at the s ame time, 'however, tits shape does not apparen t ly di ffer from that common in Germa n specimens . The marg in of the orbital cavity produces a ho ri zon tal shelf in the fore par t. The dorsal surface of the postorbital is striped by the groove line of the postorbital and a section of the infraorbi tal line, b ranching in a m ode characteristic of the Pachyosteidae. The suborbital has a narrow and elongated for e part, basall y limiting the orbital cavity w hich is provid ed with a mesial lis t only indicated by an Inconspicuous r ib. To the hind the suborbital broad ens out in to a "blad e" . The posterior m argln of the orbital cavity is p rovided with a list [placed somewhat lower than the remain ing outer su rface of the bone. In the Wild ungen specimens this lis t is n ot always dis cernib le and has n ot been shown in the reconstruction of Gross (1932b, fig. 4A), owing to w hich the posterior part of the suborbital appears here narrower than it is actually. Its presence, howev er, in on e of th e Wildumgen specimens, h as been quite d oubtlessly ascertained by the presen t writer, together w ith the essen tial resemblance of the sh ape of the suborbital in the Polishand German specimens, The specimens from Galezice suggest that the p osterior :blad e of the sub orbital w as n ot so nearly honizontalas has been shown in the reconstruction by Gross, and that it d id not give the impression of being so much elon gated owing t o its longer contact with the postsuborbital . The infraorbital line running over the outer surface of the anterior, narrow portion of the suborbital meets the forward p art of the "sup rarrnaxill a

AL B

/---­ / / I I I I I I

Fig. 12. - Pachyosteus bulla: A reconstruction of head skeleton, side view; B diagrammatical drawing in the fore part of the ventral armor and its position in r elation to the head sh ield (above-dorsal surface, below-ventral surface). AL anterior lateral, AMV anterior median ventral, C central, IL interolateral, IG infragnathal, AVL anterior ventrolateral, M marginal, N nuchal, PM postmarginal, PN paranuchal, PrO preorbital, PSO postsuborbital, PtO postorbital, SO sub- orbital. Stippled a rea - supposed position of the gill slit. UPPER DEVONIAN FISHES 325

The m argin in contact w ith the p ananuchal and the m arginal, umf or­ tunately damaged, app aren tly seems to h ave Iorrned several ben ts. The lower margin, in contact with tJh e postsuborbital, about 23 mm long, h ad :: rectilinear course, meeting the free posterior margin of the element at a sharp angle. Owing to this and to the r ela tively great length as well as to the position of the suborbital, the posterior margin of the head sh ield did n ot project obliquely d ownw ard 'begdnning fro m the point of connection between the joint and the shoulder girdle. On the lev el of contact between the post marginal and the postsuborbital, it formed a bent to the rear, to agree with the corresponding embayme nt of the forward margin of the 'anterior lateral. The hind m argin of the post­ marginal, about 30 mm long, is provided with a list fitting, with the h ead lowered, into the cleft of the forward margin of the anterior lateral. The infragnathals, both right and left, are damaged ' in their hind portions. The preserved part of the more complete left plate is 106 mm long and 18 mm broad in the back. In the front. vat a d istance of 45 mm from the anterior end, the lower margin of the infragnathal displays a gentle curve limiting th e functional pact from the posterior "blad e" . The upper margin displays a marrow surface (4 mm), characteristic of genus Pachyosteus, covered by minute denticles arranged at random. This surface, somewhat raised in relation to the upper margin of the p osterior blade of the jaw, g rows na r rower to the front and back. At th e hack it occupies the upper edge and partially overlaps the m esial surface of the infragnathal. To the front it passes onto the outer sur­ face of the element. On the mesial surface of the fore part of the infragnathal, star ting at the bent on the lower margin mentioned above, a crest extends tow ard the fr ont limiting the low er and upp er parts of the m esi al surface. Not far from the antenior end of the bon e, this crest is arcuately directed to the u pp er margin of the infragnathal, a t the same time becoming flatten ed ou t and fr om the top limiting the excavation in the symphyseal part . Ven trally this excavation is concealed by an ex tr emely fine, probably perichondral, osseous lame lla. The anterior supragnal.1h at is a very small bone, not exceed ing 12 mrn in length. It cons titutes an a r ched bl ade, w jth a finely tubercu­ lated surface on the lower m argin, sim ilarly as in the infr agnathal. On th e upper margin of the 'anterior supragnathal, at approxim ately its mid­ length is placed the upper process. Hence the whole elemen t h as a triangular outline with the lower margin convex and the other two concave. The posterior supragnathal, contrary to the above considered element, is very long but equally narrow. Its length attains 52 mm, which is approximately the 'length of the functional !part of the infragnathal. The 326 JULIAN KULCZYCKI

thickness and th e h eight of the element do n ot seem to exceed 3 mm. Backwar d and forward this jaw grows narrower and mor e pointed. Over a considerable central portion of the lower margin there is a n arrow cuspidate surface.First it occupies the lower margin. and a p art of the outer sur face of the element, and forward overlaps jJhe mesial surface, following the shape of th e Jnfragnethal surface. The parasphenoid (fig. 1:3) shows a different structure than that in Dinichthys, Heintzichthys, Malerosteus and Brachythoraci gen. inde t., described in 1956, and apparent'ly represents h er e a mor e p rimitive

S.tr:, ,, 1'>Ii!~~,...:.:.<.Ii'"

F ig. 13.- Pachyosteus bulla, the p arasphenoid: A dorsal view, B side view, C ve ntral view ; tn hypophyseal depression, m. gr unp aired cen- tr al groove, s. tr. transver se groove . evolutionary stage. This is a !bone, 25 mm in length and 14 mm in maximu m width. Its smooth an d plane ventral surface has the outline of a somewh at elongate a nd r ounded h exagoo. Anteriorly it is provided with a process wh ose later al m argins fi rst rum parallel and then con­ verge at a nearly r ight angle. Dorsally, in the hin d pa r t of the par a­ sphen oid is a raised area with an extensive bowl-like, pituit ary dep r ession (fig. 13, fh). On both lateral s ides the margin of the d ep ression is particularly raised. To the front of this d epression, on its outer em­ bankment extends a transverse groove (s.tr.) whose lateral ends are directed arcuately back ward . Medially this groove branches off into an unpaired ramification (m. gr.) direc ted str aigh t forward, along the central line, over the dorsal surface of the anterior process , which is here somewh at r aised. These grooves probably contained in te r nal carotids fu sing here into one unp aired canal entering the cranial cavity. The anterior lateral (pI. IX, fig. 2) is not completely preserved. The hind part and a considerable portion of the lower margin are missing. The hind part of the anterior lateral was an elevated blade (over 45 mm in height); forward it graduaIly elongated into a long and relatively narrow process. The fore margin of the bl ade, w hich, at t he back, limited the cleft separating the shoulder girdle fro m the head shield, was UPPER DEVONIAN FISHES bifur ca ted in to two lists bounding a slit-Iake depression, The m esial list is wider and continues farther forward. Downward it grows narrower while the sli t bec omes shallower. A rather small oval tubercle is placed at th e lower end of the slit and inward of the external b ase of th e list. The top of a triangular embayement extending fro m the fr on t m argin of the anter ior lateral is dir ect ed toward this tubercle. In that point the margin of th e considered ele me nt bents contin uing in the upper margin of the anter ior p rocess. Immediately in fron t of the avail tub ercle and bene ath th e m en tioned embayment this m argin forms a semicircular elevation and then is directed arcuately forward, wid ening ou t lanceo­ lately. The anterior process of the anter ior lateral is 67 mm long. First it is relatively high (over 15 mm on the level of the semicircular eleva­ tion) and blade-like with thickened upper margin only. Forward it grows narrower so that its terrninat end, with 'a length of 16 mm, lis a narrow rod. This Ipar t of the infcagnathal was in contact with the ioterolatera1. The interolateral is an elongated rod-like 'bone. Unfortunately, it :5 not well preserved. It was about 40 mm long with thickness and width not exeeding 3 mm. It was in corrtact with the 'ant erior margins of the anterior ventrolateral and the anterior median ventral. Following the shap e of the ventral armor, the interolaterals of both s ides met at the front at an angle of 75°. The anterior median ventral is lim an extremely fragmentary sta te of preservation, but on evidence of the preserved remains it seems to b e of a broadly rhomboidal shape. The an terior ventrolaterals completely con cealed its hind part from the ventral sid e. The width of th e anterior median ventral was about 45 mm, with the length probably at nearly th e sam e figure. The ante rior ven trola teral is also lin an unsatisfactory s tate of p re­ servation. It w as 35 mm w ide and approx, 60 mm long. A detached medi an dorsal, ref erable to a cons iderably la rger indi­ vi dua l, was found wi th in the same Ga lezice beds. The shape of this specimen suggests that it was a rep resentative of P achyosteidae. Sin ce Pachyosteus bulla is thus fa r the only known represent-ative of this family, recorded from the mentioned dep osits, it is probably referable to that sam e species. The wid th of the median dorsal is 144 mm , w ith the m axim um leng th at nearl y th e same or slightly higher figure. Alon g the cen tr al line of th e ven tral surface extends a low 'cres t, whose m aximum heigh t probably did not exceed 40 mm, measured fr om the outer surface of the m edian dorsal: In its lower part, i.e. 12 mrn below the base, the carinal process has a spoon-like depression but slightly protrudes (about 15 mm) 328 JULIAN KULCZYCKI beyond the hind margin of the element. As a whole the median dorsa is almost flat, ve ry gently arched . The hind margin is somewhat round ed. The same beds have also y ielded a fragment of the head shield of a large individual, comprising a consid erable p art of the postorbital and the iparanuchal, as well as s mall fragments of the p reorbita l and the central. The preorbital, 40 mm wide and over 70 mm long, is traversed by the central canal which bends an gularly an d continues as a canal running on the marginal. The upper portion of the postorbital part o f the infraorbital canal branches off from that angular bend in a m anner characteristic of the Pachyosteidae. The bad state of preservation renders impossible a more precise description of the remaining elements. Suggestions as regards the dimen sions of the studi ed specimen m ay be supplied by the size of the area, here measuring 94 mm, between the ramification of the sensory line grooves on the preorbital an d a similar poi nt on the joint part of the p aranuchal. The r emaining specimens consist of the infragnathals of small individuals, also re-corded from the Clymenia beds of Galezice, Remarks. - The species Pachyosteus bulla, established by Jaekel in 1903, is thus far the only known representative of genus Pachyosteus. Gross (1932b, p . 16) characterizes it as follows: "A large and broad pineal wedges deeply between large centrals and at the front comes in contact with an equally broad and blunt r ostral. The marginal does not unite with the central, The nuchal is broad and short... The occip ital canal is indicated on the par anucbal only a'S a vestigi al d etail. The op ercular can al is present. The r amification of the p ostorbital canal is characterized by the approac h of theinfraorbital to the postorbitalcanal. The infraorbital canal overlaps the marginal without uniting with the prolongation on ·th e subor bital. Lack of tube r cles on the head shiel d and th e body a rm or ". The specim ens from the Holy Cross Mts. fully coincide wi th this description. Neither did a close cornparaeive stu dy of the here described material with the Wildungen sp ecimens fr om the coll ections of the Geol.­ Paleontological Institute of the Humboldt University in Berlin reveal an y cardinal difference. More conspicuous lateral angles at the fron t of the head shield in some of the German specimens are associated with a post-mortem deformation, while certain differences in proportions of the r ostral dimensions are within the limits of individual variations shown by the Wildungen specimens. To sum up, the only difference of any significance are the relatively large dimensions displayed by some of the Galezica specimens. Since, however, th e majority of the Gal ezice specimens also in this respect agree with those from the Rhine province. it must be recognized that they are conspecific. UPPER DEVONIAN FISHES 329

Occurrence. - The distr ibu tion of Pachyosteus bulla, thus fa r only recorded from Manticoceras bed s of th e Frasnian in the Rhine provinc e, must now be extended thr oughout the F amennian of the Ho ly Cross Mts .

Gen us Ox yosteu s Jaekel {1911) Oxyosteus SIp. (pl . VIII, f ig. 4 a-b) In the material recovered from horizon II (Mid dle Prasnian) of th e Wietrznia hill a fragment of th e median dors al was found characteristi­ cally bent medialy, so that the two sides meet at a r ather sharp angle. Ow ing to the damaged condition of the surface it is hard to ascertain wheth er the bend in the b on e produc ed here a shanp edge 0 '1' a rounded r idge. The ventral surface sh ows the fore portion of the carinal p r ocess whi ch w as low and bro ad. The outer surface ds fairly uniformly tuber­ cu la ted (tubercles up to 1 m m in d iameter), On areas from which the outer bone layers have been eroded the tubercles are very minute, suggesting that their size depended on the dimensio ns and the ag e of the individual. This m ed ian dorsal r esembles the correspond ing element in r epresentatives of genus Ox yosteus. Two species of 'th is genus differing dn character of ornamentation are known from the Frasnian of the Rhin e 'Pro vince. In respect to t he size of tubercl es the Wietrznia speci men comes close to Ox yosteus magnus Gross, from which it differs in the density and regularity of tuberculation resembling mo re that in Ox uosteus rostratus Gross. The median dorsal from the Holy Cross Mts , is considerably larger than the Ger man specimens, attaining a length of over 90 mm and a width of 40 mm, while these dimensions in specimens of O. rostratus and O. m agnus, as described by Gross, are 35 X30 m m and 54 X35 mm respectively.Hence it is quite likely that we are dealing here with a very large individual of O. rostratus, though it is not quite inadmissible that it may be a new sp ecies. (Pachyoste u s bulla, as shown above, m ay also a ttain cons ider ably larger dimensions than those display ed by the material from the Rhine p rovince).

Gen us Holonem a Newberry, 1889 rad iatum (Rohon in coll.) Ob rucev , 1932 (pi. X. fig. 2 a-b) ]932. Holonema radiatum Obr ucev; D. W. Obrucev , Holonemidae.... p. 100-1 07, pl. 5, fig . 2- 5; pl. 6, fig. 5; pI. 7, fig. 3. Matel·ial. .-.:.... A fragment of the anterior dorsolateral, by Gorizdro­ Kulczycka ide ntifie d as Holonema radiatum, fr om the collect ions of the Geological Institute. JUL~ KULCZYCKI

Description. - In ShCliPE' 'and dimensions this specimen is almost: identical with that element described by Obr uchev (Obrucev, 1932, fig. 7-8). The anterior margin is characterized by a similar indentation backward at the base of the joint process and by a d oming beneath that process. The lower margin of the free area in the fore part shows an upward curving at a small distance from the anterior margin of the bone. The incomplete and damaged state of preservation of the studied specimen does not permit any precise measuring of the angle of inclina­ tion of the upper and lower parts of the plate, in relation to the horizontal plane, as indicated by the ax is of condyle and by the course of the lateral lines. This angle seems likely to be somewhat greater than in the specimen described by Obruchev. It should be noted that these angles must have been slightly different in relation to the anterior and posterior margins of the bone. The condyle is large, conical, growing narrower toward the rather sharp mesial end. Itis in the shape of a high pyramid with a triangular base, toward the top more rounded in section . One edge of the base corresponds to the anterior edge of the condyle which does net extend here beyond the anterior margin of the anterior dorsolateral. The wall of the condyle base opposite to this edge passes without sharp limitation onto the inner surface of the element, not producing a thickening directed backward, as m 1y be seen for example in Dinichthys. The lower edge of the base continued downward as a thickened crest extending mesially along the broad anterior margin. On the outer side of this margin, under the condyle, is placed the characterist:c subgleno.dal \pro'cess (Obrucev, 1932) . The anterior edge of condyle meets the lateral surface of the bone at an angle of about 125 0 (in the plane indicated by the axis of condyle and by the horizontal branch of the lateral line), suggesting that in this region the body broadened out backward. The free outer area of the anterior dorsolateral has a characteristic ornamentation consisting of costae produced by the fusion of 1-2 rows of minute tubercles. In costae made up of two rows of tubercles, the tubercles are of smaller dimensions. Similarly as in the ElP ecimen described by Obruchev, above th e h orizontal lateral line, these costae are directed from the front and botton upward and backward. Over a small distance under the mentioned line they are disp osed parallel to its course, while farther downward they are placed normally to the lower margin of the free area. Besides ornamentation the outer surface displays the already men­ tioned groove of the sensory line which projects backward from the base of the condyle. At a distance of about 35 mm from the anterior margin of the base, a ramification branches off from this canal at' UPPER DEVONIAN FISHES 3:tl

a nearly right angle, first directed upward and then curving in an arch ba ckward. The course of this canal in our specimen differs from that figured by Obruchev only in a less conspicuous backward and then for­ ward curve in the fore part of the canal and in the longer backward ex­ tension. Nevertheless, in the reconstruction published by Oaruchev (1932 , fig. 26) the canal likewise projects far to the hind. The preserved fragment is 80 mm in size, as measured along the anter ior margin, and 75 mm as measured along the h orizontal lateral line; the thickness of the bone at the base of th e condyle attains 25 m m, and an average of 7 mm in other 'Por tions. The length of the condyle measured on the anterior edge is 55 mm. There are ab out 10 cos tae on a distance of 10 mm. Occurrence. - According to Obruchev (1932) Holonema radiatu m is reco-rded from the Frasnian deposits within the Leningrad and Donetz basins. At present its distribution must be extended into the Frasruian of the Holy Cross Mts.

Genus Deueonemas n. gen. Diagnosis. - The same as for the species described below, the only one so far known.

Deveonema obrucevi !J '11 . sp,

Material. - A fragment of the median dorsal. Diagnosis. - A rather small representative of th e Holonerruidae, wi th the outer surface of dermal bones ornam ented by small, detached tubercles (only occasionally 2-3 being fused together), on the whole dispersed at random , but in so me parts, as for ins tance 'in the mesi al portion of the median dorsal , arr anged in longitudinal rows. Holo tupe. - A fragm ent of the median dorsal (1. G. ). Description.-A small fragment of the median dorsal, represen ting not more than one third of the en tire bone. Nevertheless it may be as certain­ ed on the preserved part that this ele men t displayed the elonga tion and roof-like doming common in Holonemidae. Both lateral p arts descend from the edge projecting on the central line sideward and downward, and meeting at an angle of 135°. On the ven tral side a smooth area is visible, narrow and elongate, limited by straight nearly parallel margins an d on the central line provided with a slightly indicated, shallow and rather broad groove. Sid ew ard from th e smoo th area a surface ex tends

8 Th e generic name is composed of th e initials of Prof. D. W. Obruchev. o This sp ecies is dedicated to Prof. D. W. Obruchev of th e Paleontologlcal Institute of the Academ y of Sciences of the USSR. 332 JULIAN KULCZYCKI on each side, overlapping the posterior dorsolateral. They form bands, un iformly perforated by p ores, wi th wid th constant throughout the entire length. The out er dorsal surface is covered by rather small tuberc les, d ispe rsed at rando m and only along the ridg e in the cen tral line arranged in to lon gitudinal row's, sli ghtly converging to the rear, without, howeve r, fusi ng toge th er. The ro ws of tubercles over a small area along the lateral m argin of the bone are less di stinc tly m arked. In the ar ea of convergence of th e longi tudinal rows or tubercles, which corresponds to the ossification centre placed n ear the hind end of the m edian dorsal, th e tubercles a re som ewhat smaller, ar ranged more dens ely and more at random. Th e fusion of two or three tubercles is obs ervable at very few p oints of the bone.surtace oruly ; the axes of the t h us formed thicke­ nings do not display any d efinite direction. The width of the m edian dorsal us 60 mm; the width 01 the free ventral area is 25 mm, that of the central groove is 6 mm; the depth of the groove UiP to 2 mm. The tubercles are with an average diameter of 1 rnrn . Remarks. - 'I'he particular genera of the Holonernidae family have been differentiated on character of ornamentation. Their typical repre­ sentatives, Holonema, Gyroplacosteus and lVI egaloplax, are distingutshed by the presence on the outer surface of the bone of variously shaped costae, produced by the fusion of tubercles. Ornamentation consisting of detached tubercles is common in genus Aspidichthys inclusive of " Aspidichthys" ingens, by Schmidt (1938) re-named Anomalichthys ingens, whose appurtenance to the Hol onemidae w as established by Gross (1937). The median dorsal of th e last form, however, differs rather strongly in shape from that d isplayed by typical Holonemidae with which more closely allied is Deveonema obrucevi n . 'SIp. The systematic position of the inadequately known American forms, referred to genus Aspidichthys, is not thus faIT quite clear, moreover they differ from the here considered Holy Cross Mountains specimen in the character of ornamentation. Hence the latter form may not be associated either with the American forms of genus Aspidichthys or with A. ingens from the Rhine province. On the whole, in brachythoracids, neither the character nor the very presence of ornam entation , is 'a generic feature, nevertheless, it seems most likely that this is not s o in Holonemidae which constitute a rather peculiar stock line of these fishes. Therefore, the present writer accepts the suggestions of Prof. D. W. Obruchev, an authority on Holonemidae, kindly communicated by letter, to separate the above described form into a new genus. Occurrence. - Horizon II of the Frasnian in the Wietrznia hi II in Kielce. UPPER DEVONIAN F ISHES 3Ct3

Genus A nomalichthys (Koenen , 1883) Schmidt, 1938 Anomalichthys ingens (Koenen, 1883) (pI. XI, fig. 1- 3)

1883. Aspidichthys ingens Koenen; A . Koenen, Beitrag zur Kenntniss . 1895. Aspidichthys ingens Koenen; A. Koenen, tiber einige Fischreste .. 1932b. Aspidichthys ingens Gross; W. Gross, Die Arthrodira..., p . 47. 1933b. Aspidichthys inyens Gross; W. Gross, Die Wirbaltiere..., p. 48-49, fig. 14A ; pI. 6, fig. 5; pI. 7, fig 2. 1937. Aspidichthys ingens Gross; W. Gross, Ibld., p. 40, fig. 20; pI. 5, fig . 4. 3938. Anomalichthys i ngens Schmidt; H. Schmidt, Uber Aspidichthys ..., p. 313-317.

Material. - Fossil remains of this form occur in great ab unda nce in the Holy Cross Mts, and m ore particularly so in the hills of Wietrznia, Psie Gorki and Kadzielnia within the town of Kielce where they ar e confin ed to the upper p art of the F rasnian d ep os its, i.e. to the Mantico­ ceras beds. In t'he majority of cases they are recovered as fragments of markedly large and thick bones, with characteristic verry large tubercles on the surface. From the collections of J. Czarnocki two spe­ cimens only permit closer classification. They are a large fragment of the hind part of the median dorsal (pl. XI , fig. 1, 2) and a part of the nuchal. Description. - The preserved rruchal fragment comprises the de­ pression by Heintz (1932) called "double s ockets", !probably also the whole fore part of the element. The mentioned depression shows here no signs of bi-partition and is sharply limited by steep enba nkmen t at the hin d par t only. To the front it forms the continuation of the nuchal depression produced by its r oof-like d oming. The median ridge is clearly indica ted on the outer side, from this ridge the surface descends sideward forming planes inclined at an angle of 110°. The en tire outer surface is covered with rounded, smooth, blunt tubercles of large dimensions, occasionally fused together, with diameter from 3 to 6 rnm and height fr om 1.5 to 3 mm, sometimes finely radiating at the base. The thickness of bon es, least at the front in the central li~e where it is 7 mm, gradually grows to 20 mm in the posterior lateral area and attains over 55 mm at the hind of the "double sockets". The length of the fragmen t measured in the central line on the outer surface is 90 mm. The anterior border bears traces of weathering. It seems likely, however, that this ele ment did no t project much farther forward and that is was outlined like a low triangle, similarly as is the cas e with Hol onema radiatum in Obruchev's re construction (1932). The other fragmen t bel onging to this form represents a large part, i.e. approximately one t hird, of the m edian d orsal The margins of the elemen t have not been preserved in any (par t of the specimen, yet it m ay JULIAN KULCZYCKI be supposed that this bone wa s gen tly rounded posteriorly, like that in the specimen described by Gross (1937). The roof-like d om ing of t he median dorsal is well indicated by the r idge which is particularly strong in its m ed ian part, al so b y the lateral planes which d escend to the Slid e 0 at an angle of 120 • The carinal process is preserved complete at the anterior end of the element. Like in other Holonemidae it is in the shape of an dnconspicu ous con e. From other represen ta tives of this family, Anomalichthys ingens differs in the cons iderable robustness and relative shortness of this process , w hile the diameter of its base is quite lar ge. It is also charac ter ized by greater inclination. The hind surface of the process is striped by a longitudinal groove and p laced very nearly normally in rel ation to the outer surface of the median dorsal. The outer surface of the bone is ornamented w ith the same kind of tubercles as those on the above described nuchal rplate. The only noteworthy feature is that, posteriorly, around the ossification centre, wh ere the tubercles are markedly smaller, they are arrangcd jn fairly distinct concen tric rows. The length of the preserved fragment, without the carinal process, is 284 rnrn at the central line, the maximum width 257 mm, the thickness up to 36 mm; the height of the median dorsal 115 mm (measured from the tcrp of the carinal process normally to the plane. of the median ridge on the dorsal side). The length of the carinal p rocess is 70 mm: the depth and width of the groove on that process is equally 15 mm. Remarks. - In 1883 Koenen described fr-agmentary fossil 'rem ains of huge brachythoracids from Upper Devonian deposits (Manticoceras beds) of the Rhine province. He referred one of them to genus Aspi­ dichthys Newberry, recorded from North America, esta:blishing the new specles A. ingens to dn clude it. The r emaining specimens were by h im differentiated under the new generic and specific n ame of Anomalichthys scaber. Specimens r epresenting these two forms differed. solely in size of tubercles ornamenting their bone surface. In the opinion of Gross (193'2 b) this may have been due to the difference in age of the particular individuals. As has been proved by Gross (1937), all the other specimens sub­ sequen tl y described under the name of ArlOmalichthys belong to genus Brachydirus, hence the name of Anomalichthys has become void of meanmg. In this connection Schmidt (1938) suggests to retain the generic name of Anomalichthys for the large-tubercles forms from the Rhine province, since their appurtenance 1'0 the i nadequately described Ame­ rican genus Aspidichthys is doubtful. Schmidt's opinion as to the differentiation of the American and German large-tubercles forms seems reasonably correct. though the UPPER DEVONIAN FISHES 3;}5 suggestion of usin g fOT Eu ropean forms the generic name of AnomaUch­ thys, formerly ap plied to quite a different ge nus, does not seem a very happy one. Nevertheless, in order to avoid further misunderstandigs the present writer here adopts Sch mid t's proposition . The following br ief characteristics may be added to the information pub lished by Koenen (1883, 1895) and Gross (1932 b) on this species : to Anom alichthys in gens belo ng huge brachy thoracids wi th bones orriarnen­ ted by large tubercles up to 10 mm in diameter. In th e central p or tions of the b ones the tubercles are small er an d arranged in m ore or less distinctly concentric rows. The med ia n dorsal, as is usual in Holonomidae, is elongated and roof-domed, with the ri dge in the central line . The caninal process is in the sh arpe of a mass ive, stout cone and. is m ere in cli ned downward than lin oth er Holon emidae. The nucha l is rel atively shor t and bears a ven trall y dep ression, not limited anteriorly and without bi-partition. Occurrence. - The Frasnian (beds with Manticocems intumescens) f.rom the. Rhine province and the Hol y Cross Mts.

Genus Opercnauosteus'" n. gen.

Diagnosis. - Brachythoracids with dermal bones ornamented by basally sharply limited meandering ribs and by less numerous, fl attened tubercles in the median part of bones. From Gyroplacost eus the men­ tioned genus differs in a more delicate p attern of ornamentation, less regular arrangement of el evations and their fl attened tops.

Operchallosteus vialowi 11 n. sp , (pi. XII, fig. 3) Material. - A fragmen t of a posterior ventrolater al. Diagnosis. - The sam e as that for the genus given above. Holotype. - A fragment of th e ri gh t posterior ventrolateral (1. G.) (pl. XII, fig. 3). Description. - The preserved fragment, 145 mm in length and 85 mm in m aximum width, probably represents a major part of the posterior ventrolateral. It is distinguished by extremely characteristic ornamentation, at first s ight resembling the relief on the bones of Botbriolepidae, It, however, does not exhibit' rniorostructure characteri­ stic of . Its ornamentation somewhat resembles that of genus

10 The generic name is an allusion to the beautiful ornamentation from the Greek word in:wxui.A:r,O, meaning unusually beautiful. It This species is dedicated to Prof. O. S. Vialow from Lvov. 336 JULIAN KULCZYCKI

Gyroplacoste us . In the latter for m, however, contrary to Operchallosteus, the details of ornamentation are much more d ensely arranged, parbicularly so in the centr al parts of the bones.Moreover, in Gyroplacosteus, the tubercles are d om ed and p la ced near the r ibs throughou t the su rface of bon es, while in Operchallosteus they are bluntly truncated an d concentrated in the mi ddl e portions, the !per ipheral part being occupied by meandric ribs here fus ed together into an in tri cate meshwork. Occurrence. - The F rasnian, horizon II (median) in the Wiatrznia hill of Ki elc e.

Infraor do Arctolepida Within Upper Devonian dep osits of the Holy Cross Mts., particularly so in the F amen nian (Cheiloceras bed s) of the K adzieln ia hill, there is an abundance of placoderms r emains or namented by tube rcles, sometimes ar ranged in concentric rows and displaying a str uct ure common in Arctolepida. Unfortunat ely all the found specimens are very fragmentary and specifically indeterminat e. On accou nt, however, of their copiousness th e writer feels justified in mentioning th eir occu rrence.

Ordo Anti archi Genus Bo thriolepis Eichwald, 1840 sp. (p i. X II, ·fi g. 1)

In h er paper 0'11 d ipnoan fishes (1950) from the Devonian of the Holy Cross Mts.Z. Gorizdro-Kulczycka mentioned the presence of Bothriolepis cf. maxima and B. pand eri within Mi ddle Frasn ian deposits. T he mater ial at the pr esen t w riter's disposal con tains but one fragmen t of bone refer ab le to a large representative of gen us Bothriolepis (pl . XII, fig. 1) w hose specific position it is n ow difficult to establis h .

Pl acodermi incorti ordinis (p i. X II, fig. 4)

Fragment of an armor pla te recovered from the Cl ym enia beds of Ga­ lezice is a markedly interesting though puzzling specimen. Its ou ter surface is ornamented by peculiar wrinkles whose symmetric arrange ment sug­ ges ts that we are here dealin g wi th an unpaired elemen t of the m edian series. The bone structure in dica tes that this form d id not belong either to Antiarchi or to Euarth rodira. Lack of exact infor mation with respect UPPER DEVONI AN FIS HES to bone structure of other placoderme hinders 'proper comparative studies. The outer surface of the bone consists of a paper-thin layer wrinkled as mentioned a bove and covered by tubercles hard ly discernible with the naked eye. The tubercles display the absence of "u ndipolar " cells so characteristic in Euar throdira, Layers of bone and bone cells are here disposed parallel to the outer surface. The remaining mass of the element is made up of a spongy bone tissue with large intra-trabecular spaces. The bon e is up to 4 mm in thickness. In sp ite of the meagreness of the available evidence this form undoubtedly represents a n ew p lacoderm.

REMARKS ON CERTAIN STRUCTURAL DETAILS OF SOME BRACHYTHORACLDS The posterior sup ragnathads described in the preceding chapter, though distinguished by extremely divergent trend of specialization, pro­ ducing in the former a crushing jaw apparatus and a cutting one in the latter, both display the same essential structural scheme. This may, .moreover , be encountered in a number of other representatives of Bra­ chythoraci, as for instance in Coccosteidae and Mylostomidae, in dicating that the various types of jaw apparatus in these forms have all evolved from the same initial type, illustrated by the catching jaw apparatus in primitive Coccos teidae. A specification of the poster ior supragnathals belonging to variou s representatives of .Coccosteus and Plourdosteus is given in Heintz's paper (1938). In these genera the p oster-ior supragnathal is shaped like an elongated plate. On its upper margin, somewhat in front of the point corresponding to its mi dlength, we fin d the upper process which is the place of at tach men t for ligamen ts or muscles . From 'it's top toward the lower border, a thicken ed r idge exte nds on the mesial su rface. On the outer lateral side, the an terior slightly inflate d por tion forms a lateral process whose vertical edge is provided with a row of denticles. A simi­ lar ro w of de n ticles is also noted on the hind edge of the posterior supra­ gnathal. Observations made by Gross (1932b) and by the present writer in­ dica te close resemblance in th e structure of the jaw apparatus between Coccosteus and forms allied with it, on the one hand, and Pholuiosteus, on the other hand. In the latter the pos terior supragnatha l was also provided w ith a lateral process. The chief difference consists in the far ad vanced atrophy of th e rows of den ticles in Ph olidosteus, although the sa me tendency may also be observed in Coccosteus and Plourdosteus (Heintz, 1938). An additional point of difference in the jaw apparatus of these brachythoracids is th e d ev elopment of a conspicuous cuspidate process on the fore end of th e infragnathal 'in Pholidosteus:

Acta Palaeont ologtca Polontca - Vol. II/4 JULIAN KULCZYCKI

With fu rther reference to th e new representatives of brachythoracids described in this paper, it sho uld be no ted that in T'omaiost eus we ag ain encounter all the str uctural elements of the posterior supragna thal common in Coccost eus and similar forms. Th e cardinal d ifference h ere is again that of th e absence of vertical rows of den ticles. The remaining differen ces are merely those of proportions, pa rticularly respecting the poorer d evelopmen t of the upper and lateral processes , and the h igher but stout er vertical thickening on th e m esial surface of Tomaiosteus. In genus Malerosteu s the posterior supragnathal, besides the absence of vertic al rows of dentides, is characterized by a conspicuous thkik€r.: \" ng of th e elemen t and stron ger development of th e crushing surface replac ing the lower edge. On thi s surface two tubercles are visible, one of which is placed near the corner corresponding to the vertical thickening 0111 th e mesial s urfa ce, while th e other one is near the tolP of the lateral 'process. In genus Mylostoma, with an extremely vspecialized crushin g tyip e of jaw ap par atus, we note an unusually strong development of the lateral pro cess and relatively shortene d anterior portion of the element. At the same time the whole posterior supragnathal becomes flattened out dorsa-ventrally. As in Malerosteus, th ere are here 'also two tubercles on the lower crushing surface. In Mylostema, however, the outer tubercle, here centrally placed, is much more robust. Toward the top of the lateral process, an inconspicuous ridge or crest extends from the outer tubercle. It might be supposed that this tubercle corresponds to a similar product in Malerosteus. In Dinomylostoma, which is a less strongly specialized form regarded as ancestral of genus Mylostoma, there is only one tubercle on the crushing surface. For the present, these structures cannot be definitely homologized since it may reasonably enough be conjectured that we have here a case of convergence. In any case, the jaw apparatus in Malerosteus and Mylostcmidae displays the same trend of specializa­ tion; in Mylostoma, however, this process has attained the highest' evolutionary stage, whrle in Malerosteus 'it is less advanced, similarly as in Dinomylostoma. In both the latter genera the p oste rior supragnathal shows the strongest adaptation for th e cr ushing of food, the infragnathal somewhat less so, while th e anterior supragnathal has retained its catching character. In Mylostoma, however, also this element is tho­ roughly adapte d to the crushing of food . This functional differentiation of the jaw elements in brachythoracids r esembles the differentiation of teeth in high er where the fo re part cfdentition retains th e 'Catchin g character, while the hind p art is ad apted to some manner of fr agm ention of the food . A striking feat ure is that of the high er crush ing specialization of the upper e lemen ts which generally develop a larger fu nc tional surface than is th e UPPERDEVONlAN FISHES

case in the lower ones. This applies to the posterior supragnathal in brachythoracids, and seemingly to the upper molars of p enissodactyls and ruminants. Naturally, in all the quoted examp les this specializabion d eve lops quite independently and is probably referable to greater Iixedness of the upper jaw element and greater freedom for movement in the lower ones. Thus, as on the one hand, Malerosteus a nd Mylostornidae display a similar trend of sp ecialization, so on th e other hand, Tomuiosteus an d ' th e dinichthyids adapt a similar mode of taki ng their food too . Of the latter forms it is the dinichthyids which attain a higher degree of sp ecializati on . Their adaptation to th e cutting of food p asses fr om an ~xt re me narrowing of the posterior supr agnat hal to a complete disappear­ ~ n ce of the lateral process. . The adaptive similarities do not indicate close gen etic: alliance of t he consider ed forms. 'Tomaiosteus and the dinichthyids are probably rela ted by distant coccost eidal ancestors and undoubtedly represent dis tin ct evolu tionary lines. Neither is lit 'in adm issible, as mention ed above, that s imilarities in the jaw apparatus of Mylostomidae and Malerosteus are th e result of convergence, but it is not quite ou t of the ques tion that th ey are the representatives of the same s tock. In Pachyoste us, sim ilarly as in the jaw u!ppa~atus of din ichthyids, the posterior supragnathal lack the la teral process.H ere, howe ver, this lack may be a primary phenomenon, as is also tJhe lack of the differenti ation of denticles and of arrangement into well marked r ows in the catchin g jaw apparatus of this genus. In this respect, Coccosteus and Plourdosteus, provided w it h distinct rows of denticles, have probably attained a higher degree of sp ecialization. The structu re of denticl es in Pach yosteus is of p articular in ter est. Even an outer insp ection made under slight m agnification r eveals a d if­ ferent d egree of the union of d enticles wdth the jaw surface. Some of them h ave fused complet ely w ith the underlying' bone, wh ile others are separated from it 'by a slit on the peripher al portion of dts base. ThE' re lati ve ind ependence of the d enticles is also clearly shown lin cross sections of jaws longitudinally cutting through the denticle. In suc h a s ec tion (fig. 14) we dis tinctly see the upper edge of th e jaw with a cap-like d enticle seat'ed on it (d) . The peripheral p or-tion here is en tirely fused w ith the underlying bon e, While in the centre it is sep arated by a slit. In the d en t icle itself we may differenti ate the b ase. the lateral wall and the internal cavity. In the wall (fig. 14 C) there are three zones, indistinctly limited. The outermost 'and thinnest layer c onsists of an osseous tissue with typical b one cell-spaces connected b y 'branch ing canaliculi and arranged more or leSIS parallel to the p lane JULLAN KULCZYC~

Fig. 14. - P ach'!Jost eus bulla: A cross section of upper m argin of the infragnathal with the denticle (X 20), B longitudinal section of denticle (X 150), C section of denticle wall (X 300), Cp pulp cavity, D dentine, 0 osseous tissue, U interme­ diate layer, d denticle. UPPER DEVONIAN FISHES 34'1

'of the tooth. This layer, thin at tlhe apex of the tooth, growsrthieker downward 'Passing into the osseous tissue of th e b ase (fig. 14 B). IV10re centrally, in the lateral tooth w all 'w e note the intermedia te layer (U) with elongated cell vsp aces and with the long er axis placed normally .to the plane of tooth. Among the canaliculi. for the cell processes one, branching from the central extremity of the cell-space, is sli ghtly more conspicuous, particularly when seen under small magndfication, owing to whi ch these cell-spaces resemble the "unilPolar" cells of Heintz and Gross, so characteristic of the tubercles or den ticles present in the jaw and dermal bones of Coccosteus and Plourdosteus. The inner, thickest layer (D) consists of a tissue by th e writer regarded as m odified dentine. In this layer, from the inner cavity w hich may thus be considered as pulp cavity (Cp), canaliculi r adiate, thicker mesially and n arrowing p eripherally. The \peripheries of many of these canalic uli are connected with the mesial pole of the cell -spaces in the intermediary zone. This, together with the more clearly expre sse d canaliculus branching from the outward pole of the cell-space imparts the impression that the cell-spaces lie on the cours e of the radial canaliculi. The cell-spaces on the intermediar y zone, as stressed abo ve, differ from the rema ining bone cell-spaces by their arrangeme nt only.We might thus be dealing with dentine canaliculi connected to canaliculi of bone cell proc esses, which is quite a common occurrence. The on ly pecu liar feature is th at the network produced by the connection of minute ramifications bran ching off throughout the course of the radial canaliculi only slightly differs from that formed by canaliculi for the bone cell processes, with which it is also connected without delimitation. The connections between den tine canaliculi, however, are also observable within the dentine of other fishes, while their appearance is greatly diversified. Moreover, the dentine ap pears to be of a transitory nature leading to complete disappear­ .ance and replacement by an osseous tissue , whence may result its non­ .typical pattern. This problem will be discussed later. Another peculiar feature of denticles in Pa zh-qosteus is that th eir outer p ar t consists of bone tissue. In thds respect denticl es 'in th e s tudied form resemble tubercles in the dermal bones of Plourdosteus livonicus {Gross, 1933 a) and of Holoptychius (Bystrow, 1942). When comparing the structure of denticles on the infr agn ath al and that of tubercles on bones in Plourdosteus livonicus, Gross (1933a) concluded that between them there is no essential difference. They diff er only ,il11 the predom in ance of dentine in the jaw denticles and that of "unipolar " cells in the tubercles of dermal bon es. At that time the jaw denticle'S were by Gross considered to be d ifferentia ted tubercles. Evidence con tained in B ysbrow's pap er (1942), howevervshows a reversed 34:2 JULIAN KULCZYCKI state of conditions in Crossopterygii, d.e. that the tubercles of th e dermal bones are modified denticles . In coincid ence with those s tudies and with observati ons of Gorizdro-Kulczyck a (1953), during the evolution of Crosso­ p te rygii and the Dipnoi the or iginally uniform d entine coating was desintegrated into detach ed denticles. On t he dermal bones the latter were subseq uently transformed into osseous tuber cles and finally disap­ peared . The d esin tegration of the d entine coating must have occurred at an earlier date, w it hin the mouth cavity, where it ceased without producing the complete d is appearance of dentin e but w as fol lowed by the next devel opmen t and differen tiation of denticles. The present writer concurs with the views of 0rvig (l957) and Gross (1957) th at t he disappearance of d entine may have taken Iplace also in th e Arthrodira. The difference consists in the ea r lier occurrence of the process in the placoderrns, and, in its final stage, that of the transforma­ tion of denticles dnto osseous tubercles and th eir subsequen t disappea­ rance, having involved not the dermal bone denticles OID ly. but th e denticles within the mouth cavity as well. Such an evolutionary t rend of this proc ess cis suggested in the brachythoracids both b y the gradual disappearance of ornamentation on the d ermal bones during phyl ogenetic evoluti on and by the fact that the presence of mandibular denticles is, in the first place, common in primitive forms of the brachythoracids, while in forms occ upying a higher phylogenetic stage they usually disappear. During the ontogeny of some forms (Heintz, 1938) young individuals may still be provided with denticles lacking in older individuals, but this fact can also be referable to the wearing off. The dinichthyids m ay be mentioned among those s tocks in which the disappearance of denticles is observable in the course of evolution, In the prdmitive Diniehthys herzeri, the jaws are provided with a row of denticles on the edge, whose histological structure is unfortunately unknown. Also in the primitive D. pustulosus there are vestigial denticles which disappear completely in more advanced representatives. Neither are traces of dentine ascertained in their jaws (Heintz, 1931 b), as is also th e case in the posterior supragnathal of Tomaiosteus. Here this element is made up entirely of an osseous tissue with densely arranged osteons separated by diminutive inter-osteonal trabecules and having relatively n arrow vasoular canals. The or ientat ion of these canals is in most' cases p erpendicubar to the cutting edge. In Malerosteus too, nearly the whole infr agnath al is made Uip of the osseous tissue. Here, however, there is no suc h r egular arrangement of the osteons and the inter-osteonal t rabecu les are strongly developed. It is only in the centrad part of the tooth;-like cusp, placed at 1Jheantenior end of the rnfragnathal that it UPPER ' DEVONIAN FISHES

.w as pcesible to ascertain th e presence of a detached den tine system . .In the dra wing (fig. 15) we see the 'cross section of the lP ulp canal (Cp) en circled by concentric den tine layers with the characteristic thin dentine canaliculi. P eripherally (U ) there are few, de t ach edc e ll ~siPac es. ThE' dentine system .is on all sides surrou nded by osseous tissue (0) whose

o u o Fig. 15. Malerosteus gorizdroae n. gen. n, sp., fragmentary cross section of tooth-lik e process of the infragnathal (X 150); CP pulp canal, 0 osseous tissue. U transitory layer, D dentine.

-cells w ere united with the processes of the odontoblasts, Th e absen ce -of any traces of resorption aJ11d the character of the union of the dentine system witlh tlhe surrounding osseous tissue suggest that we are not -dealing here with a tooth of the older generation, functional at some e arlier date, an d now enclosed in th e bone. Most likely it' is the vestigial r emains of teeth once existing in ancestral forms of Malerosteus. The presence of cell-spaces and their connection with dentinal cana­ licu li in th e dentine system of Malerosteus resemble the picture of the tissue differentiated by 0rvig (1951) under the name of "semidentine". The difference between the dentine as figured in 0rvig's drawing (1951, fig. 2a) and that seen in Malerosteus consists in that the latter form ·diSlPlays cell-spaces grouped in some peripheral par ts of the dent ine .system on ly. This system attains here considerably larger dimension s th at' those in Plourdosteus canad ensis investigated by 0 rvig (1951). In th e latter species the dentine seems to icor reqpond with the tissue only of th e per ipheral part of th e 'd en tine syste m in Malerosteus. It may be JUL~ KULCZYCKI r easonably inferred that semidentine is n ot a distinct type of d en tine. but a trans ition in respect to both Us position, in termediary between b-one tissue and typ ical dentine , and to the process of the disappearance of dentine and its substitution by the bone tissu e. In other words, at the mo ment of deposition of the first den tine layer 'in the spaces betw een the already existing bone foundation (it is known that in teeth the dentine is laid downInward) the odontoblasts here are not fully · d~ff eT ­ en tia ted since they display certain osteoblas tic features and, in some cases , are enclosed within the dentine mass. It is not before the nex t s tage - which is no t a ttain ed by Plou rdosteus oW1iTIg to the small dimensions of its dentine eleme nts, but observable in genus Malero ­ steus - tha t the differen tiation of odontoblasts is completed . Hence they persist under their ty pical form. 'I'hese d isturban ces in the d iffe­ rentiation of ·odontoblas ts are most likely associated with the general process of the disappear ance of dentine and its subs titution by bone tissue, experienced by brachythoracids and probably by other Ar throdira too. These irre­ gularities may some time s have been very in­ tensified, lead ing gradually to less and less ty ­ pical form of cells w ith the app earance and behaviour in termed iate between odontoblasts and ost eoblas ts. An d thus, on a bon e fragment, foun d in the Famennian shales of the Kadziel­ nia hill, mar e clos ely inde te r minate but pro­ bably belonging to Arctolepida, the present writer was able to ascer tain all the transitions from typical odontoblasts to osteocytes (fig. 1 6. ) ~ These transitory cells resembled to the so-called unipol ar cells. They are bottle-shap ed and pla­ ced normally to the pl ane of tu bercles. The thick cell process, starting Irom the ou tward Fi g. 16. - Fragm entary pole, narrowed into a neck-like shape, is next sectio n of tubercle on the divided into sma ller cell processes connected bone of Arctol r p ida g vn, et sp. indet. Iro-n the F 1­ wi th th e network formed by the cell processes m ennian of Kad zi lni a, of normal osteocytes. Some of the here desert­ showing ' changes in the shape an d arrange m mt of bed bottle-shaped cells were con ne cted by the cell sp ac es fr om th ~ v as­ outward pole with the canaliculi radiatin g from cul ar canal towards the outer surface of bone. the outward part of vascular canals. Others, retaining the ch aracteristic bottle-shap ed ou t­ lin e an d the disposition of bones normal to the surface, produce processes throughout their surface, as is usually the case in ordinary osteocytes. As was kindly communica ted to the writer in a letter, Prof. W. Gross UPPER DEVONIAN FISHES 345 has also enc oun tered most diverse passage forms ranging from unipola r cells to osteocytes. The disappearance of dentine from the jaws of brachythoracids is not reas onably clear insomuch that it involves not only forms taking their fuod without fragmentation, but those with the cutting (as in Dinizh­ Lhys and T omaiosteus) or crushing (Malerosteus) type of jaw apparatus too. This phenomenon may perhaps be interpreted by the inability to cease of the process of dentine disappearance, in spite of modified food requirements,' hence the transformation IProC€s>s of dentine or its substitution by bone cannot be here reversible. Another interpretation is that of the development of structures talking over the function of dentine, such as the prog ressin g keratinization of tlhe epithelium covering the jaws. A possibility of this kind was already suggested by Newberry (1889) in connection with the low er jaw structure in Ti tanichthys. At the presen t moment it is s car cely p ossible to suppor t this hypo­ th esis. Most au thors assert the p resence of distinct t races of the wearing off in jaws of brachythoracids. The material ava ilable to the present writer provides no decisive evidence in favour of either of the two alterna tives. The character of the cutting edge 'in th e p osterior supra­ gnathal of Tomaiosteus speaks rather against the wearing off of jaws. In Malerosteus the wearing off is quite admissible. Somewhat confusing is th e lack of structural differences in th e surficial part of the tooth-like infragnat hal cuE\p between the side contacting the anterior supragnathal, w here the abrasion should take place, and the m esial , free part where it could not take place. It should also be noted that in its normal position the tooth-like in fragn athal cusp does not extend to the top of the corresponding notch on the anterior supragnathal and that in order to m ak e a notch it must have turned r ound its axis, this being quite improbable. Another supplementary remark is that keratinization of the epithe­ lium, which may explain the disappearance of dentine from the jaws of brachythoracids and at the same time also contribute to the clarificatlon of the prob lem of the growth of these elements in brachy­ thoracids provided with a crushing or cutting jaw apparatus, is riot a strange though indeed not a very common occurrence among fishes. A compensation in the growth of jaws on the basal side does not seem very likely, since, e.g. this would alter the position of the functional part of the infragnathal in relation to its hind blade, making it subject to continuous structural modifications parallel to resorption of its upper edge.In a contrary case, the upper edge of the functional part, as it -wears off, would h ave gradually to lower its position in relation to the upper edge of the h ind portion of the mandi ble. T his phenomen on 346 JULIAN KULCZYCKI actually occurs in P tyctodon tidae. They are p rovided, however, with. strongly developed dentin e and the reasonable supposition is that the wearing off w as not so very r apid h ere. It m us t have been much mor e in tense in brachythor acids whi ch lacked den ti ne structures" unless w e take dnto consideran on th e p ossibility that their food consisted of soft substan ces only.In this case, how ever, th e p owerful ad ap tations to cr ushing and cutting would be strikingly ou t of proportion with the softness of food . These difficulties in the interpretation of the function and grow th of jaws dn brach y thoracids, ad ap ted for crushing and cu tting food , do not con cern forms with a catching jaw apparatus, as for instance Coccost.eus. In this genus the wearing off of denticles (and in older individuals ­ even that of functional edges of jaws) may have occurred. Still less do­ th ese difficulties con cern with such forms as Pa chyosteus, in which the wearmg off of jaws did not certainly take pl ace and where the growth of jaws did not necessa rily differ from that in the r emaining dermal bones. With growth, the denticles, which in Pachyosteus were probably inserted entirely in the soft tissue of the mucuous lining, mo st likely experienced r esorption and w ere substituted by the next genera­ tions of denticles formed b y indep endent d en tal p ap ill ae subsequently fusing with the jaw. Besides the peculiar structure of its denticles, Pachyosteus is an interesting form by the shape of its parasphenoid too. A descri ption of this element is given i n the former chapter of this :paper; here it is only noted that the shape of this eleme nt in Pachyosteus seems more primitive than that in dinichthyids an d even in Buciuinosteus, In Pachyo­ steus, similarly as in most Arctol epida, the internal carotids pierced the cranial base sid eward from the parasphenoid and rim on the dorsal su rfac e of the latter. In op pos ition to Dinichthys they would not produce here a transverse ariastornosis , but would fu se together into one unpaired vessel entering the cranial cavity. Their behaviour must probably have been similar in Malerosteus wh ose parasphenoid was descr ibed by th e present writer in a short p ap er of 1956 as an elem ent: of a new genus of Brachythoraci. The diversity in the structure of the p arasphenoid in various brachy­ thoracids as well as the behaviour of the above mentioned vessels will, in the future, probably serve as additional evidence for the differentia­ tion of separate evolutionary limes within this stock. During the investigation of brachythoracid material from the Ho ly Cross Mts. the writer's attention was d rawn to the shape of the hin d margin of the head which, beginning from the joint socket to as far as the boundary between the ipostmargtiJnal and the postsub orbital, is UPPER DEVONIAN FISHES 347' usually provided with a list som ew hat lowered down in relation to th e remaining surface of the bone. This detail, together with the shape of m argin on the anterior lateral, indica tes that with 1Jhe head lower ed the above part of the hind margin of the h ead a rmo r extended beneath the fron t margin of the body ar mor. A t the same time, there are lists and tuber cles on the edge of the antenior lateral (on e of them placed at the bend of the free anterior lateral edge seems to b e rather constant), which must have served as a ttachment points for the lig aments and muscles. All this indicates that the gill slit could not have pr ojected above the boundary between the postmarginal and the postsuborbitai. It must have, therefore, b een placed ventrally, as supposed by Heintz (1932). Such a p osition of the gin slit lis also suggested by the structure of S ynauchenia, one of the Wildungen forms, in which the dermal head armor bones were ankylose d with those of the thoracic armor , thus fill in g up the cleft b etw een these regions. Since Synauchenia belongs to brachythoracids extremely fl attened Iaterally and was certainly a necto­ nic form, it is h ar dly ad missible that a secondary displacement removed its gill slits onto the ventral side, this being common rather in ben thon ic forms . Ther efor e, we ought to accept that the ventral p osition of the gill slits ]n S ynauchenia corresponds to an ancestral character with structure mor e typically b rachytnoracid. Reverting to the p roblem of growth, those of the jaw elements excepted , the w riter wish es to draw attention to a specimen of posterior med ian ventral of Malerosteus gorizdroae?, described in the present iP3!Per (pl .III, fig. 3). As stated above, this specimen shows two growth lines, of which one consists of the margin of the free area. The lack of apposition of new bone layers on the outer surface, has here made possib le the preservation of the or-igina'l outline of a young stage of this element on the bon e of an older individual. Bein g not of frequent occurrence in t he brachythoracids, it probably resulted from a distur­ bance of the normal growth process. The disturbing factor, to judge from the normal peripheral growth, acted uniformly th r oughout and did not penetrate to any great depth of the bon e. In the writer's opinion it may have been the iPreSSUTe exercised by the bottom, on the ventral surface of that rather benthon ic . The consider ed specimen also d isp lays traumatic dam ages. They consist of two d amaged areas on the right and left sides of the e lement. The tubercles present on the bottom of these a reas, which might be d amaged by the jaws of some aggressor, ind ica te that the ind ivid ual to w hi ch the s tudied element belo nged , escaped death and that th e regenerat.ing process had already started. Disturbances in the apposition of new bone layers on the outer 348 JUL~ KULCZYC~ surface are also observable on th e nuchal of Malerosteus gorizdroae?, where near the ossifi cabion centre we find an area without the last bone layer. It is only encroach ed by detached tubercles differring in size from tubercles of the preceding generabion. This produces s omething lilke Westoll's line. Whether we a re d ealing h ere with a physiological phenom enon or with the consequence of a d isease of p eriosteum of the bone itself, is d ifficult to conj ecture. Sinc e the lack of continuity do es not affect the las t layer only, the writer is inclined to th e latter COIrl­ jecture. It should be stressed that th e picture presented by the ossification centre in the nuchal of Malerosteus gorizdroae?, i.e. the presence of detached tubercles of the new generation on the surface of the preceding layer, in some brachythoracids results from physiological growth.To say in Dinichthys d. tuberculatus (pl. VIII, fig . 4), amo ng tubercles of the older generation, intimately connected with the corresponding continuous layer, are isolated and seemingly a~posed tu bercles of a n ew generation, or r ather generations. At firs t they are flat and lens-like, but with growth they become hi gh , cupola-like or conical. It would ap pear that the polymorphic tub erculation, so character istic of this f OI1m , is due to that tY'P€ of the growth process. As is observable on specimens of M. gorizdroae?, th e tubercles become larger with gr ow th, h ence also with age of the individual. This also holds good in the specimen of M. gorizdroae where, in parts with re­ moved last bone layer (e.g. in the cou rse of preparation), the disclosed earlier layer displays fin er tuberculation . In the mentioned cas e the tubercles dricrease with th e growth of the individua l. In Dinich thys pustulosus the behaviour of tuberculatccn is different. Alth ou gh it has not been h ere possible to ascertain the ornamentation on deeper bone layers,

REMARKS ON THE S YSTEMATICS OF BRACHYTHORACI Th e structure of th e parasphenoid, tog ether with the structu re and character of the mandibular denticles in gen us Pachyosteus exclude the descent of P achyosteidae from the Coccosteidae. These are cer tainly two distinct stocks. Their independence has already been stressed by S tensio (1944) in the course of studies Gin the pectoral girdle. As a result of these studies Stensio separated two groUIPS of the Bnachy thnraci ,' i.e. UPPER DEVONIAN FISHES 349

Coccosteornorpha and Pachyosteornorpha. The former were made to include Coccosteus, Plourdosteus and Pholidosteus, the latter the majority of the remaining forms. Whereas the opposition of Pachyosteidae to the Coccosteidae se-ems reasonable enough, it ds difficult to acceptthe natural ch ar ac ter of the Pachyosteomorpha group, It seems more likely that they comprise a number of different evolutionary lines, while some of them (dynichthyids) may perhaps be related with the coccosteids, according to the former belief. Unfortunately, the parasphenoid of the last mentioned forms is unknown. Should its structure prove to resemble that of the Dinichthyidae, it will then be p ossible to SlPealk of a line going from forms represented by (by White, 1952, shown to belong 1'0 the brachythoracids) . th rough the Coccosteidas to the Dinichthyidae. In any case, the dinich­ thyids and thcccccosteids 8IPIP ear to be equally distant from the pachyo­ steids, The latter display a number of primitive characters, such as the dental structure,' the parasphenoid and possibly also the scapulo-coracoid. But this last point does not seem quite doubtless dnasmuch as the structu­ ral peculiarities lin the fore IPaTt of this elemen t, as indicated by Stensio (1944), may as well be the result of reduction associated with the reduction of the dermal bones in fuis area. Besides these primitive cha­ racters, the Pachyosteidae display certain features of specializa tion which, in the writer's belief, comprise the disappearance of the spinals. In these r espects, however, the dinichthyids are more primitive. Unfortunately, th e writer does not know whether the scapulo-coracoids of other Pachyo­ steomorpha than Ensoosteus, mentioned by Stensio 'in hi s p aper on the Arthrodtran pectoral girdle, have been studied. Th e mentioned genus belongs to the Pachyosteidae. Owing to the meagreness of information av ailable on th e structure of th e brachythoracids, it is difficult to establish a satisfactory classi­ fication which would depict the actual genetical connections between the various families, uniformly treating all the representatives. In this state of conditions it seems preferable to admit a division of this stock in to distinct families without any hints as to their relationships, such as has been done by Romer in his handbook " Paleontology" (1947). We must note, however, that in this, handbook n ot all of these units are of the same rank. To say, the Coccosteidae, unto which Romer among others also includes the dinichthyids and several forms without a clear position, if not an artificial assemblage, would have to be consi­ dered of a higher rank as compared with the other differentiated fa­ milies. It rather seems that the majority of the latter have been ranked too high and that in the future th ey will have to be re-layed 'to th e 350 JULIAN KULCZYCKI rank of sub-families. For the time being, however, the retention of the present classification seems more to the purpose. Having here touched the problem of the sys tem atics the writer also wishes to mention the occ asional te ndency to separate the Mylos tomidae into a unit of equal r ank with the group em bracing virtually all of the r emaining Brachythoraci (Berg, 1955). Mylostomais an examp le of highest spe cialization as regards the mode of the nourishment. This spe cializati on probably in volves only the jaw apparatus, which in Br.3­ chythoraci displays great vari ab ility. It is SynaU'::henia m ar kedly m ore than Mylostornidae that deviates from the general structural scheme of the Brachythor aci, S ynauc henia is thus far the only kno w n for m which , by the fus ion of the head armor with the thoracic armor, loses on e of the m ost di agnostic features characterizing the gTOUIP to which it belongs. But ev en this apecialization, so exceptional among the Brachythoraci, does no t jus tif y a special treatment of the Symauchen iidae. This applies all the mor e so to th e Myl os tom idae.

STRATIGRAP HICALA ND GEOGRAPHICAL DISTRIBUTION OFDES CtUBED FOltMS The mat e rial desc rib ed in th e present paper h as been coll ected fr om but fou r sites, th ree of w hich are adja cent. In this connection it is d ifficult to conceive a ge neral id ea of the fish -fa una involving the entire Holy Cross Mts. area. The Upper F rasnian excepted, thechacacter 0.£ the p articular Upp er Devoni an horizons is here depicted by fi nds from various sites which m akes difficult the investigation of the evolution of the ichthyo-faunal assemblage in the progr ess of time. Another Impor tant deficiency is that the m aj or ity of forms is represented by single specimens only . The above mentioned de ficiencies in the availabl e material greatly li m it, if n ot al together excl ude, th e possibility of r eas on ably vabid bio­ stratigraphic, paleogeographic and paleoecological conclusions, and imply the necessity of ex e rcis in g the u tmost care in considering these pro­ blems. In spi te of the inadequacy of th e available material, the picture it gives us is, surely, not com pletely accidental. Hence, in this chapter, the writer will risk an atte mp t at picking ou t som e r egularities in the vertical and horizontal distribution of the brachythoracids, at least within this very limited part of the Holy Cross Mts, region. The attached table contains a specifi cation of brachythoracid forms r ecorded from the DIPper Devonian of the Holy Cross Mts. Each vertical column limited by continuous lines contains forms occurring within on e horizon, marked with a cross. UPPER DEVONIAN FISHES

As is sh own in this table, Din ichthys pu stulosu s seems to h ave th e greatest vertic al distribution, as i t ranges fro m the lowermost Frasndan upwards inclusive of the Lower Famen nian. Second to it in this respe ct is D. denisoni which embraces the whole of the Famennian period. The next to be mentioned is Ma!erosteus gorizdroae which probably lasted through the Middle and Up pe r F r asn ian. Anomalichthys ingens, recorded f.rom three sites , is amomg th e fOI1JIls re presen ted by a great er number of specimens. It is confined here to th e Upper Frasn ian, i.e . to h or izon III, or the Man tk oceras beds of Czarnecki. The Middle Frasn ian is characterized by greatest diversity of brachy­ th oracid forms. The m aj ority of forms, how ever, recorded from this horizon are represented by single specimens only. Differences of fa unalcomlPosition in the particular horizons are stressed by the sequence of for ms foll ow ed in the speciftcation ta ble. Th us it is seen that of nine form s recorded from the Middle F r asn ia n, two only were found in other h orizons . Ou t of the three Upper Fr asn ian for ms, one, characterized b y widest range- of distribution, h as not been r ecorded fro m other horizons. Two out of the four Lower Famennian forms and three out of the fou r Upper Famen n ian on es, are apparentl y characteris tic of these horizons on ly . Thus tJhe depicted p icture, 'if at all reflecting the actual faunal changes, would naturally characterize the considered area, since some forms encoun tered outside of the Holy Cr oss Mts. display there a some­ what different ver tical distribution. To say, Dini chthys tuberculatus an d D. pustulosus are know n fro m the Givetian to the Lower Famennian, Pachyost eus bulla an d genus Oxyosteus - from the Frasn ian, while Anomalichthys ingens iS3lPpa'I'ently confined to th e U1ClPer Frasnia n only. The asynchronous occurrence of the mentioned forms within diffe­ rent areas is very pr obable, but does not exclude the necessity for the introduction of a number of corr ecticns on the consid eration of facial difference. It should again be stressed here th at th e d lscussion of ei ther the vertical or horizontal distribution of Iorms known on d etached specimens only , is altoge the r impossible. Local assemblages of forms r ecorded from the particular sit es h ave been shown in th e same table, grouped in vertical columns limited by broken lines . It can be seen that this d oes not much alter the arrange­ ment of columns as compared with the preceding on e. The p ossibility of representing in this Wlay two different interrelations in one sp eciti­ cation table is dependent on tlhe circumstance that the faunas from the 352 JULIAN KULCZYCKI

Distribution of the d escribed brachythoracids

Horizon Fra sn a n Famennian

II III Cheiloce- Clymenia Low- Upper Mid- 'I ras beds beds Species e r dle with Manticoceras

I I. Pachyosteus bulla + Titanichthys ko zlowskii + Dinichthys cf. tuberculatns + Dinichthys denisoni Dinichthys cete rus + S tenosteus? sp. -+- Dini chthys pus tu Losus + + + '? + A nomalic hthys in gens + + 1- Mal erosteus gorizdroae .) + -+- Tomaiosteus grossi +

Dev eonema obrucevi -t- Ho Lonema rad :atum + O per ch allosten s vi aLowi

Oxyost eus sp. + I I Plourdost eu s sp. -I- 1- - -- Psie . . Locali ty Wietr znia Gorki Kadzielnia G al ez ice 1 UPPER DEVONIAN FISHES 353 particular horizons have been recorded - th at from the Upper Frasnian exc epted - from different sites. This is not, however , an accidental outcome only, being to a large m easure dependent on facial differentia­ tion, since, with the exception of Galezice, there are several horizons in Every exposu re and th e sear ch for mat erial w as made at every locality throughou t: th e whole section. Three sites: Wietrznia, Psie Gorki and Kadz ielnia, w ith Frasnian out crops, lie within the range of the Kielce facies which is a more shallow-wa tsn- one as ccmp ared with th e deep er Lysa G6ra facies, s ta rting at a sma ll distance to th e north and outcropping in the north-western periphery of Ki elce at Czarn6w, and at Zagorza and Radlin to th e east. The men tioned sites are situat ed on the elevation of the Frasnian sea floor connected (Czarnecki, 1947) with the p er iphery of the Dymin y anticline. In spite of the small d is tances separabing th ese localities, rather importan t facial differences are displayed in dep osits of Lower and Middle Fra snian age occurring there. To say, at Kadzielnia, Lower and Middle Fr asnian de posits are ma de up of poorl y layered reef limestones, first of the gastropod-brachiopod fauna, with Loxonema poloni cum, Pleurotomaria ka dzielniae, and higher up of the brachiopod- coral fauna. At the same time, more bitum inous and clay ey , layered limestones, with thin sha le intercalations were deposited at Wdetrznia. To the east of Wie trznia (Zagorze, Radlin) and to the north-east of Ki elce (Sluchowice, Czarnow), the reef facies disappears completely and Frasn ian deposits are represented by thin-layered limestones, marls and shales passing into similar Famennian d eposits (Czarnecki, 1947). Diff erences between the Lower and Middle Frasnian fis h faunas in Wietrznia and Kadzielnia are apparently associated with facial diff e­ rentiation. While in Kadzielnia these deposits have not, thus far, yielded any fish remains, r elatively many forms have been 'recovered from Wietrznda, Som e of them (Plourdosteus, Holonem a, Bothriolepis) are characteristic rather of continental deposits, or d eposits formed at a small distance from the land. The diversity of sedimen ts and their faunal assemblage (e.g. presen ce of Lingula banks) indicates that they are m a­ rine shallow-water dep osits. Although, it is n ot impossible that the above men ti oned pl acoderms might h av e lived in an open sea, nevertheless the ir remains m ight h av e been transpor ted by currents. The latter alternative m ay be suggested by their sporadical occurrence, since they are represen ted by single specimens. From the eventua l transport w e must, however, exclud e that of being dragged over the sea floor, since there are no traces of ab rasion. Ptyctodontidae and Dinich th ys pu stulosus, found in fa ir abundance, are certainly forms who had lived here at that time.

Act a Pal aeon t ologtca Polo n lca - Vol. II j4 3'54 JULIAN KULCZYCKI

Durin g th e Upper Frasnian , within the Wie trznia-Kadzieln ia area, the facies grows more uniform. We h ave here thick-layered limestones w ith Mantico::e ras intu m es.:: ens and Hypothyridina cu boides. A t the same time the ichthyo-fauna also becom es m ore uniform, w it h Anomali::hthys ingens as the m ost com m on r epresen tativ e occ urring in all the three localities. In the K ar czowka hill only, to the w est of K adzielnia, does the r ock y r eef lim estone facies s till persist . Tho last m en tioned fo rrn is abs en t there. F urther facial changes and mo re ex tensive unifor mity occur a t the beg in n in g of the Famenrni an when the e n tire Ki elce region is cover ed by t.hin-Iayered slate lim es tones and m arl s of d eelP sea origin.These cir­ cumstances are responsible for th e grea ter diqpersicn of fi sh r em ains. They are somewhat mo re ab und ant within the lowermost Cheilo ceras b eds only. Dinichthys pustulosu s con tinues to o ccur along w ith oth er forms which, most likely, did no t occu r during the Frasnian, at least in the con sidered area. Wi thin upper Ch eiloceras and Clymenia b eds of this r egion, fi sh remains are very sporadlcal and no brachyth cracids h ave th us fall' been recorded there. At that time (correspond'ng to the Clymenia bed s) these fishes occur in greater abundance south-west of Kielce, i.e. in Galezice (at a distance of 15 km), The Clymenia b eds r est here dir ectly on' Givetian limestones and di ffer m arkedly from those occurring within the remaining area. While there they are made up of alternating thin beds of slates, marls and Iirnestcnes, wi th a total thickness of up to so m e hundred meters, in Galezce th ey ar e reduced to a lam estone la yer, 3 to 4 rn in thickness containing an abundant and diversifie d fauna (gastropods, goniatit es, cly rneniids, solitary corals, crinoids, trilobites, as well as plant remains). Of the brachythoracids, in addition to Dini::hthys denisoni known from the Ch eilo ceras bed s of K adzielnia, T itan i::h thys k ozlowskii, Dinichthys cf. -tu berculatus and Pa::hyosteu s bulla h ave also been encountered. Though faci al features m ay also b e of significan ce h ere, it is hardly probable t:hat Pachyost eu s - s hou ld it have existed within this area already in the Frasnian, as it d id in the Rhin e province - could have remained and been discovered among the abundant br a chy thoracid r emains in Wi etr znia. This for m m ay not, therefore, have penetrated into the Holy Cross Mts, basin before the Upper Famennian time. It would, h owever, be puzzling why it did not appear at an earlier time, together with Anomalichthys ingens whose occurrence is noted within Mantkoceras beds botJh in the Rhine province and in the Holy Cross Mts. It is possible that the absence of Pachyosteus in the Upper Frasnian beds of the Kielce region resulted from different environmental requirements of this form. UPPER DEVONIAN FISHES 350

As bef ore m ention ed , Dinichthys d. tuberculatu s is

The w ide range of d ist rib uti on of D. pustulosus gives rise to the problem respecting the centr e of its evolution and migration r outes. In North America this form is known from the higher part of Mi ddle Devonian beds (Hamilton ). It's range of distributi on is limited to areas lying west of Kentucky, It is not before the Upper Devon ian time (Genesee) that D. pustulosus makes its way, together w ith Manticoceras intumescens, eastward, into the state of New Yor k. According to Eas tm an (1907 a), the migration route of this form leads from Europe, through Manitoba in Canada. In western Europe, how ever , this form has not thus far been discovered. Within the Holy Cross Mts., i.e. in central Eurcp e, its p resence is not recorded befor e the Upper Devonian time. From t he Frasnian of Russia, Ob ruceva (1956) has descr ibed, under t he na me of Dinichthys lic harevi, a form which is close ly r elated. Though th e pr esent knowled ge on D. pustu losu s is extremely inadequate and may give ri se to grave errors with 'resp ect to its d istribution, neverthe­ less, on the information now availab le it seems more likely tha t the centr e of evolution of this form was in the w es tern p ar t of North Ameri ca, or m aybe, in Siberia w hence it w as shifted equatorially on one si de fr om the w est to the eastern 'p ar t of America an d on the other side from the eas t into Europ e. Counterw ise, D. tuberculatus having developed in the eastern p ar t of North America h as m igrated at the close of th e Fame nn ian through western Europe (Belgium) to cen tral Europe (P oland).

A PPENDIX

ON SOME ICHTHYODORULITES AND ELASMOBRANCHIAN REMAINS FROM THE UPPER DEVONIAN O F THE HOLY CROSS MTS.

Genus Setitacaru ruis'? n. gen. Diagnosis. - P air ed spines of mod erate size , built of osseous tissue, provided w ith large dentine teeth, sp arsity aligned in a asymmetrical row an th e hi nd edge of the spine.

Sentacan thus ze iiclunu skae'? n. sp. (pI. XIII, fig. 3 a-b) Diagnosis. - The same as that for the genus, since it is at present the only known species.

12 The gen eri c name has been formed from the Latin word sentus, meaning spi ny. 13 This species is dedicated to Dr. M. Zelichowska, Head of the Documenta­ tion Dept. of th e Geological In stitute in Warsaw. UPPER DEVONIAN FISHES 357

Material. - Detached specimens of spines. Holotype. - An incomplete paired spine (M. Z.). Description. - The preserved fragment (base and apical part missing) is over 120 mm in length , 15 mm in width and 7 to 8 mm in thickness. It appears, therefore, to h ave been a fairly large and slender spine. At first it gently curves forward, then, nearer the top, backward. Its cross sec tion nearer to the base is elliptic with the larger d iameter equal to 15 mm and the smaller one to 7 mm.To the top it first widens out up to 18 mm, then gets narrower again and somewhat: thickened (wid th 13 m m, thi ckne ss 8 mm). Owing to the en largement of the hind margin the se ction here is more regularly oval. At the same time the front m argin tapers out. The 'spine ret ains th is shape u ntil

Genus AiienacanthusU n. gen .

Diagnosis. - Extremely large, paired and unpaired spines, built exclu­ si vely of osseous tissue.

Alienacanthus m alkowskWr. n. sp. (pl. XIII, fig. 1 a-c) Diagnosis. - The same as for the genus. Holotype. - A fragmentary paired spine. Material. - Fragments of large osseous spines are frequently yielded by Clymenia limestones from Galezice near Checiny, Some are perfectly straight (circular in cross section) and are symmetric, unpaired

14 The generic name is derived from the Latin word alienus, meaning alien. being an allusion to the d .sstrnllarlty of th is spine to other ichthyodorulites. IS This species is dedicated to Prof. St. Malkowski, former Director of the Muzeum Ziemi (Museum of the Earth) in Warsaw. 358 JULIAN KU"L,CZYCKI structures, others are gently bent and flatten ed , thus being paired.On e specimen of this p aired spine represents a fragment sufficiently larg e for its description.

Description. - The pres erved fragment is 320 rnm In size, w ith the upper and low er ends mi ssing. It is ge n tly curved both antero-poste r iorly and laterall y. In the low er p i r t it is Ilat te ned, wi th one surface nearly plane, while the other is somew hat convex. H ere bot h the fore and hind edges are sharp. The spins is 30 m m in width and 10 m m in thickness . In the centr al p art the th ickness of the spine in creases to 14 mm, while the convex (front?) edge grow s blunter and broad er. On this edge there are fairly la rg e, wi de ly spaced denticles. Toward the top the front edge tapers ou t ag ain . On the upp er end of the apecimen the wid th of the spine is 19 mm, the thickness 10 mm. The qp ine is built entirely of osseous tissue. The canals run parallel to the axis of the element, forming occasional anastomoses . One of the canals is slightly broader (1 mm in d iam eter) , it extends throughout t he spine, somewhat n ear er to the convex (den ticl ed) edg e. The oss eou s tissue h as the character of in tervasc.ular trabecules. The fin spi nes, descr ib ed above, similarly as in Se nuizanth. us, differ from all other ichthyodoruhtes know n to the present writer in that th ey are made of osse ous tissue only. It is probable that they were co­ vered by a dermal coa ting. They may not be referred either to Selachii or to Acanthodii. It seems more likely that they for med the armor of some p lacoderms, Occurr ence. - The F amenn ian CL y m eni a beds of Galezice near Checin y.

Genus Ctenacantnus Agassiz, 1837 Ctenacaniiuse sp. (pl. XIII, fig . 2,\

This form is r ep r esen ted b y two specimens. On e of them corresponds to the median part of the spine and t he other (I. G ,) - to a p art n ear er to the base. On the preserved fragments itt may be inferr ed that the spine was moderately slender, very gently curved and equally gen tly narrowing toward the top. In these features it d iff ers from D evonian species of A merica which, moreover , are character ized b y mor e densely arranged costae. To a certain exten t it r esembles Asteroptsjchius. The spin e is strongly fl at tened with the in terior occupied by a large !pulp c avity. On bo th slides of the flat hind margin tJhere are rows of minute d errticles UPPER DEVONIAN F ISHES 359 w ith th e apexes dire cted toward the base. The sharper front margin is gently round ed. The outer surface is ornamen'ted by long itud inal costae rath er closely spaced and separated by narrow grooves which do not widen out toward the base, contrary to th e arrange-men!' we fin d e.g. in Ct.? ere ctus. Three fine costae run along the front marg in. On the lateral su rface, in th e cen tral part of the specimen, there are 10 anterior thicker costae and 7 finer posterior ones .The costae bear tubercle-like, mo­ derately spa ced swellings. In this, the orna me ntation of the Kadzieln ia specimen resembles that of Ct. gracillissimus N. & W. common in th e Carboniferous (camp. Eastman, 1902, p. 86, fig. 12), but differs from it by the non-stell at e character of tubercles. Th e tubercles here are arranged rather sparsely and on differen t: levels of th e par ticul ar costae, Th e costae mult ipl y by intercala tion, being fro m the first thicker in th e anteri or part of the later al surface .Besides Ct.? erectus Koenen two more forms have been recorded from th e Devonian beds of Europe and ten­ tatively assigned t o genus Ctenacanthus. Th ey are th e Ct. qemiinslensis Gross from th e Middle Devonian, and Ct.? jaekeli Gross from the Upper Devonian. The former differs fr om the Holy Cr oss Mts. speci men in its gen eral shap e and ornam entation characterized by greater uniforrnity of costae and stronger develqpment of crests. Th e latter Ger man sp ecies of Ctenacanthus is differentiated by being more conspicuously curved and by fewer and smoo th costae. The specimen from the Holy Cross Mts. probably belongs to a new species. Since the writer has not been able to mak e closer compara­ tive studies owing to lack of prope-r literature, he can only m en tion its presence. Occurrence. Th e Fam ennian lower Cheiloceras beds in the Ka- dzielnia hill of Kielc e,

Genus Cladodu s Agassiz , 1843 Cladodus sp, J (pl. X I II, fi g. 4) It is a very small tooth with th e basal width equal to 4 mm and th e sam e height of the main cusp. Th e main cusp, fairly robust, n arrowing swiftly fr om mid-height and tapering off in to a sharp ap ex. The anterior surface, somewhat flattened, is in its lower part ornam ented by fine vertical striae, at the sides more densely spaced . The main cusp is gently s'gmoidal. On each side of the main cusp are two lateral ones, of which the outer are large, attaining half the size of the central cusp. The intermediate de nticles are qu it e tiny. The outer cus ps are relatively 360 JULIAN K ULCZYCKI slender, conical and slightly diverging. The lower margin of the crown is gently flexuous. The tooth has been recovered from the Famennian Ch eilo ceras beds of the Kadzielnia hill.

Cladodus sp. 2 (pI. XIII, fig. 5) This tooth, found in the same beds as the above , is differentiated by th e presence of but one m ain cusp, thus r esem bling Cl, conicus and Ci, urbs-uuiouici Eastman (camp. Hussako f & Bryant, 1918, pl. 44, fig . 1, 1a). From the two last named it differs, however, by lack of ornamentation. The base of th e tooth is not v ery broad, measurin g scar cely 3.5 in width. The cusp is nearly erect , with the section circular at the base. Toward the apex it becomes flattened out. Lateral edges gen tl y pointed.

Cladodus? sp. 3 (pI. X III, f ig. 6) This tooth, r ather strongly damag ed, is by the writer onl y tentatively in clud ed in genus Cladodus, from which it differs iill that it does not develop a uniform crown, the CUSlPS 'being independently placed on a massive base. The base, m easured antero....posteriorly, is 5 mm long, measured bi-later all y probably abou t 14 m m , with height about 4 mm. At the bottom it is concave. Two cu~IPS only have been p reserved, b u t it rather seems that t he tooth was sym me trical and that it h ad three cusps . The cent ral, twice as high as the lat eral ones, is strongly bent backward, fairly robust, with th e anterior surface flattened ou t and the p osterior convex, with moderately sharp side edges . It is 8 mm high. The lateral CUEIP, only h alf that height, is also i nclined to the b ack , b ut it does not display any curves and is in the shape of an erect cone. It is fairly distant from the main cusp, and like it without ornamentation. The intermediate den ticles are missing. The studied specimen h as been found in th e Clymenia beds of Ga­ lezice,

Genus Dittodus Owen , 1867 Dittodus SiP . (X III, fig. 7, 8) The too th referred to this provisional genus is of a type to whic h the equally tentative name of " Phoebodus" h as b een assigned . As shown UPPER DEVONIAN FISHES 3tH

by Hussakof and Bry an t (1918), this type of teeth is encountered al ong with those called .Dituuius" and a number of forms intermed iary between th ese ' types. Most li'< ely therefore, th ey all belong to th e dentition of th e same fishes fr om the family Pleuracanthidae. The first of the studied specimens (pl. XIII, fig. 7) is 8 mm in w idth as measured between the apexes of the outer cusps, and 5 mm in h eight. All the three cusps are approxim at ely of the same thickness, i.e. 1.5 mrn at the base. The median is perhaps slightly more slender. The outer cUSlPS are divergent. The lateral edges are moderately sharp . Under magnification, a scarcely discernibl e stri ation may be observed on the su rface of cusps. No in termediary denticles present. Out of this type of te eth so far known, those in Dit todus priscus Eastman (camp. Eastman, 1907a, pl. 1, fig. 7-8, also Huss akof & Bryant, 1918, pl. 44, fig. 3 a-b) , with stnation relativel y delicate but more conspicuous than tha t in the Kadzielnia spec imen, come nearest to the considered spe cimen. " Phoe bodus" knightianus Eastman (camp. East­ man, 1903, pl . 4, fig . 40, 40a) resembles our specimen in the robustness of its main cusps and in the lack of intermediary cus ps. Its outer cusps, how­ ever , in addition to a more circ ular section, are quite erect, and do not diverge, as is the cas e in our specimen. "Phoebodus" politu» Newberry (camp. Eastman, 1907 a, pl. 1, fig. 12, also Newberry, 1889, pl. 27, fig. 27-28) is sometim es withou t orname ntation. It also resem bles ou r specimen in th e flattened an ter ior surface of the cusps and the presenc e there of sharp lateral edges. On the figures of Newberry and Eastman, however, the CUflPS in " Phoebodus" politus are ap paren tly more slender than those in our speci men, wh ich, moreover, lacks intermediary cusp s. ,.Phoebodus" dens-nep tuni Eastman (comp. Eastman, 1903, pI. 4, fig. 39) is entirely different since it is with ma rkedly sle nder and distinctl y ribbed cusps. The sec ond specimen of Dittodus tooth (pl. XIII, fig. 8) differs fr om th e former by the more slend er and diverging cusps. It is 5 mm in width at the base and 3 mm in height. Both specimens have been yielded by the Ch eiloceras beds of Ka­ dzielriia. Paleozoological Laboratory of the Polish Academy of Scien ces Warszawa, September 1957 REFERENCES BERG L. S. H155. Sistema r yboobra zn ych i ryb nyne zivuscich i iskopaem yc h Tr. Zoo!. Inst ., 20, 1-286 . Len in grad. BOGOLUBOFF N. M. 1916. About some bones of th e Heterostius carapace. Cleol. V estn., 265-269 . Petrograd. BRANSON E. B. & MEHL M. G. 1931. F ishes of the J efferson formation of Uta h . J . Ceo!., 39, 509-53,1. Chicago. 362 JULIAN KULCZYCKI

BRYANT W. 19129. Fossil fi sh es from the Hamilton shales of New York. A new Coccosteus from th > P ortage shales of w este r n New York. Bull. N. Y. State Mus., 281, 37-4{j. Albany. BYSTROW A . P. 19a9. z.a nustruktur del' Crossopteryg ier. Acta Zoo!., 20, 283-3,38. Stockholm. - 19412 . Deckknochen u nd Zahne del' Osteolepis un d Diptarus. I bi dem , 23. 263-289. CLARKE J . M.18S5. On the higher Devonian fau na s of Ontario County. Bull. U. S . G eo!. Surv ., 16, 9-80 . Washington, - 1903. Catalogue of th e type sp ac .mens of Palaeo zoic fossils in the Ne w York State Museum. BuLL N. Y. St at e Mus., 65, 8. Albany. CLAYPOLE E. W. 18912 a. The head of Dinichthys. Amer. G eologist, 10, 19'9-207. 1c92b. A n ew g .ga ntic Placod erm from Oh io. Ibi dem, 10, 1-4. 1893a. The three great fo ssil plac oderm s of Oh io. I bidem , 12, 89-99. · 18930. A new Coccosteus, Coccosteu s cuya h ogae. Ibidem , 11, 167-17 1. 1893c. On three new speci es of Di nichthys. Ibidem , 12, 27'5- 279. 1894. On a new Plac od erm, Brontichthys clarki from the Clevela nd shale. I bi dem , 14, 379-380 . 1895. The great Devonia n pl ac oderrns of Oh io. G eol. M ag., 1, 2, 473-474. London. 1896a. The a ncestry of th e upper Dev oni an placoderrns of Ohio. A mer. Ge ologi st, 17, 349-360 . 1896b. A new Titanichthys. Ibidem , 17, 166-169. 1E96c. Dinichthys prentis-clarki. I bidem , 18, 199- 201. 1897. A new D in .chthys,D inich thys k aple r i . I oidem, 19, 322-324. COPE E. D. 18912. On some new a nd little known P alaeo zoic vertebrates. Proc. Amer. Ph,!. So c. Phi/ad., 30, 2211 - 2127. Philad elp hi a . CZARNOCKI J . 19,35. Przsglad str atygrafii i paleog eografii dewonu dolnego Gor Sw letokrzyskich. Spruw . P . Inst. G eol. , 8, 129-200. W arszawa. 1947. Przewodn.k XX zjazdu Polskieg o Tow arzys twa Geologiczn ego w Goracn Swletokrzysk ich w .1947 r. Roc zn . P . T. G eol. (Ann. Soc. G eol. de Pologn e), 17, 2,)7-2 99. Warsz awa. DEAN B. IS93a. On Tr ac hosteus a nd Mylos toma, notes on th eir structural cha ­ r acter s. T ra ns. N. Y. Acad. Sci., 12, 70-71. New York. 18930. Contr.bution to the a na tom v of Dinichthys. I bi d em , 12, .187-188 . 1896. On the vertebral column, fins, an d ventral ar m or ing of Dinichtn ys . Ioidqm, 15, 157-163 . .1897. Notes on the ventr al armoring of Din ichthys. I bi dem , 16, '57- 61. 1901. P alaeontological notes. I : On two new Arthrod ira from Cleveland shala of Ohio. II: On the charact er s of Mylostoma Newberry. III:F urth er notes on the r el a tionships of the Arthrognati. M em. N. Y. A cad. Sci., 2. 85-123. New York. 1909. Studies on fo ssil fishes. M em. Amer. Mus. Nat. R ist., 9, 268-287. 191M. Note on the Oh io P lacoderm Dinichthys terrelli. Sc ience, 34, 801. New York. DENISON R. H: 1956. A revi ew of the h abitat of the earliest vertebrates. Fieldianu. 11, 8. Chicago. DUNK LE D.H . 1947. A new genus and sp ecies of arthrodiran fish fr om the upp er Devonian Clevela nd shale. Sci. P Ub!. Cl ev. Mus. Nat. H ist., 8, 10, 103- U , . Clevelan d. UPPER DEVONIAN FISHES

DUNKLE D. H. & BUNGART P. A -, 1939. A new Arthrodire from th e Cleveland shal e formation. Ibidem , 8, 2, 13- 28. Clev eland. - &- 1940. On one the least know n of the Cleveland shale Arthrodira. Ibidem, 8, 3, 22-47. - &- 1942a. The infer o-gnathal plates of Tit anichthys. Ibidzm , 8, 4, 49-59. - &- 194'2b. A ne w genus and spe cies of Arthrodira fr om the Cleveland shale. I bi dem ," 8, 6, 65-71. -&- 194.i3. Comments on Diplogna thus mirabilis Newb. Ibidem , 8, 7, 73-84. -&- 1945a. Preliminary notice of a r emarkable Ar th r odiran gn a thal pl ate. I bi dem . 8, 9, 97- 102. - &- 194'5b . A new Arthrodiran fish fr om th e Upper Dev on ia n Ohio shales. I b idi?l1l, 8, 8, 85-95. - &- 1946. The a nte r osupe rognatha l of Gorgonichthys. Amer .M us. N ovit., 1316, 1-1 0. New York. EASTMAN C. R. 1896a. Observa ti ons on the sh ield in th e dynichthyids. Amer. C e%gi st, 18, 222 -22 3. lSR6b. Itev iew of "On the ve rtebral col um n, fins and ve ntral arrno r ing of d in ichthyid s" . by Bashford Dean. Ibidem , 18, 3.1 6-31 7. 18D6c. Preliminary n ot e on the rel at 'ons of certain bony plates in th e dyn ich th yids . Amer. J. Sci., 4, 2, 46-50. New Ha ven. 1897. On the r el ati ons of certain plates in th e Din ichthyi dae w ith d cscrlption of new sp eci es . B ul l. Mu S. Camp. Z ooI., 31, 19-41. Cam br idge, Mass. 1898. S ome new points in d in ichth y id osteo logy. P ro c. A me r . .Assoc . Adv. Sci., 41, 371 -37'2. 1900. Einige neue Notizen tib er devonische Fis ch reste a us der Eifel. Zbl. Min · etc., 177-178. S tuttgart. 1902. Som e Ca rbon iferous Ces tra cion an d Acanthodian sharks. Bull. Mus. Camp. Zool., 39, 3, Cambridge,Mass. 1£03. Ca r boniferous fishes from th e centr al western states. I bi dem , 39, 7. ,1906a. Structure and r elation of Mylostoma. Ibidem , 50, 1-30. 1906b. Dipnoa n affinities of Ar thr od ir es. Amer . J. Sci. , ·1, 21, 131-1 43. New Haven. 1907a. Dev oni c fishes of th e New York formations. Me m. N . Y . Stat z Mus.. 10. 1- 193. Albany. 1907b. Mylostomid dentition. Bull. M us. Camp. Z ool. , 50, 209 -228. Cambridge, Ma ss. GROSS W. 193Qa. Ein wildungar Ar thr odire in Nord-Amer.ka. Palaeont. Ztschr., 14, 46-48. Berlin. 1932b. Di e Arthrodira W ildungens . C eol. Palaeont. A bh., N. F. , 19, 1. 1-61. Jena. 1933u. Di e Fische de s baltischen Devons. Palaeontograph:ca , 79 A, 1-74. Stuttgart. 19i33b. Di e Wirbeltiere des rhein ischen Devons. Abh. Preuss. L andesan st ., N. F. , 154, 1-83. Berlin. 1934a. Ub er Allolepis lon gicornis und Brachydirus sca ber . Cbl. M in. etc ., 13 , 5, 2C3 2-2 33. Stuttgart. l lli34b. Der histologische Aufbau des Phyllolepiden P anzer s. Ibidem , 12, 528-53C3. 1937 . Di e Wirbeltiere des Rh einischen Devons. A.bh. Preuss. C eol. Landesan st. , N. F., 176, 1-82. Berlin. 364 JULIAN KULCZYCKI

GROSS W. 1008. Ub er das Sp inale un d d ie angrenzenden Knochen der Brachythoraci. N. J b. M in. e tc ., 79 B, 403-4018 , Leipzig. 1938. R hachiosteus pterygiatus n . gen .. n. sp . Deche n ia na. Verh. N at. V el' . Rh ei n l. W estf ., 97 A, 183- 208. Bon n. ,1 940. Ac anthod ier und Plac od ermen a us Heterostius-Schichten Estlands un d Lettlands. Ann. Soc. Reb. Na t . Univ. T art., 46, 1-90. Tartu. Hl41. Die Borthr iole pis-Arten der Cellulosa-Mergel Lettlands. K. Sven sk l! V et . AkacL Handl., ser. 3, 19, 5, 1-79 .S tock holm . 1950. Di e p ala on tologi sche und s lr a tigr aph ische Bedeutung der Wirbellier­ fa une n des Ol d 'Reds u nd der marinen altpala ozoischen Schichtcn. Abh . deutsch. Akad. W i ss., m at. n a t. Kl., 1, 1-130. Berlin. 1957. Mund za h rie und Hautzahne d er Aca nth odl er und Arthrodiren. Paia e­ ontographi ca, A, 109, 1 ,1-40. Stuttgart. GOODRICH E. S. 1909.l1er tebr ata Craniata. 111 : Lankester, R , Trea tise on Zo ology, Pt. 9, 258-265. L ondon . GORIZDRO-KULCZYCKA Z. 1950. Dwudyszn e r yby dewonsk ie z Gor Swieto­ krzy skich (Les Dipneustes devon iens du Massif de S-te Croix). Acta Geol. Pol., 1, 1,53- 105. Warszawa. 1953. Dwudyszn e r ybv de wons k ie z Gor Swietokr-z yskich. Uz upe ln ienie. (Leo' Dipneustcs d evoni ens d u Massif d e S-te Croix . S upplemen t). I bidem, ;~ , I. 153-170 ; Consp . 39-44 . W arszawa, H EINTZ A. 1928. Ei nige Berner k u ngen tib er dar P anzerbau bei Homostius u nd H eterostius. Skrif t . No rs k e Vid. A kad., m at. na t . Kl., 1, .1- 12. Oslo. 192'9. Di e do wntonisc hen und devo nis che n Vertebraten v on Spitsb erge n , II. S kriit, Sv albard l sh av et , 32, 1-81. Oslo. 1930. E in e neue Rekonstruktion vo n Heterosti us As m . Sitzber. Dorpot.. N tit ur], G es., 36, 3/4, 1-7. T artu. ,1931a. Revision of the str uct ur e of Coccoste us decip iens Ag , Norsk; Geol. T i dskr., 12, 292-311'2. Osl o. 1931b. Unter suchungen uber den Bau de r Arthrodira. Acta Zool. , 12, 225-239. Stock holm . 193,1c. A reco nstr uction of Stenognath us e: oul d i Ne w ber ry. Ann. Mag. N ut . H i st ., 10, 8, 24'2- 249. Lo nd on . 19316. T he antero-late ral pla te in T itanichth ys. I bidem, 8, 206- 212. 1S3,l e. A new reconstruction of Dinich thys. A m er Mus Novit.; ·157, 1-'5. Ne w York. 1932. The structure of Dmichthys, a co ntribution to our knowled ge of t he Arthrodira. Dashf. Daun Me m . V ol. Archaic Fishes , ·1, 115- 2:24. Ne w Yo rk. 1034. Revisi on of the Estonian Arthrodira, I : Family IIom ostiidae J k l. Eesti L ood ust. Archiiv, ser. 1, 10, 4, 1-115 . 1935. Holonema -reste a us dern Devon Sp itzb ergens. N crsi: ::ieol. Tidssk r. , 15. 115- 121. Osl o. 1938. No tes on Ar throdira. I bidem , 18, 1-2 7. H ILLS E.S. 1936. On certain en docr a n ia l str uctures in Coccosteus . Geo!. M ag. , ;3. 2H3 -2Q6.London . H ITCHCOCKC. H . 1868. Uber Din ichthys herzer i H itch., ei ne ne ue amer ik a ni sche Form fossilcr Fische aus der Devon. N. Jb. Min . etc ., 7, p. 874. Stuttgart. H OLMGREENN . 1942.S tudies on the head of fis hes. Acta Z ool., 23, 129-26 1. Stockholm . UPPER D EVONIAN F ISHES

HUSSA K OF L. ,1905a. No tes on the Dev on ian "P la coder m"Dinich thys intermedius Newb. B u !l. Amer . Mus. Nat . R i st., 21, 27-:36. New York. 1905b . On the structure of two impe rfectly known din ichthyids. Ibi dem , 21 409 -414. 1906. Studies on the Arthro dira. Me m. Amer. Mus. Nat. Rist., 9, 103-.154. New York. 1909. The sy ste matic relationshi ps of cer tain Amer ican Ar throd ires. Bu!t. Amer. Mus. Nat. !list., 26, 263-272 . New Yor k. 1911. Notes on som e upper Devoni an Arthrodira fr om Oh io, USA. in the British Museum. Geo!. M ag., 5, 8, ,123- 128. Lo ndon. 1912. Notes on Devonic fishe s fr om Scaumen ac Bay. B u l l. N. Y. State M us.• 158, 127-139. New York. 19,13. Descr.pti on of four new palaeozoic fis he s fro m North America. Hu ll. Amer. Mus. N at . Hist. , 32, p. 245. New York. 1942. Fishes fr om the Devonian of Arizona. Amer. Mus. Novi t., 1186, 1-9. New York. HUSSAKOF L. & BRYANT W. L. 1918. Catalog of fossil fishes in the Museum of the Buffalo Society of Natur al Sciences. B :tll. Buff. Soc. N at . Sci., 12, 1-341. Buffalo ~l AEKEL O. 1902. tiber Coccosteus und die Beurtheilung dar Placodermen. Sitzber. Ges , Natur], Freunde, 103-115. Berlin. 1903. Uber die Organisation und systematische Stellung dar Asterolepiden. Ztschr. deutsch. geol.' Ges. , 55, 41-60. Berlin. 1904. Neue Wirbeltiere in Oberdevon von Wildungen. Ibidem, 56, 159-167. 1906. Neue Wirbeltiere aus dam Devon Wildungen. Sitzber. Ges. Naiuri. Freunde, 73-85. Berlin. 1907. Uber Pholidosteus, no v. gen., und die Mundbildung und die Kerper ­ form der Placoderrnen. Ibidem , 1711-186. 19,19. Die Mundbildung der Placodermen . Ibidem, 73-110. 1927. Untersuchungen tib er die Fischfauna von Wildungen. Pulaeont. Ztschr.• !I, 1/3, p. 3,29. Stuttgart. J ARVIKE. 1954. On th e visceral sk eleton in Eu sthenopteron wi th a discussion of th e parasphen oid and palatoquadrate in fishes. K. Svenska Ve t. Akad. Handl., ser. 4, 5, 1, 1-104. Stockholm . KA YSER H. E. ,1880. Ub er Dinich th ys( ?) eifeliensis von Ger olstein. Z tsch r . deuts ch . geol. G es., 32, 817-818. Berlin. KOENEN A. 1876. Ube r Coccosteus (Br achydi rus) bicken sis. I bi dem , 28. 667-66ll. 1880. Uber Cocco st eus Arten aus dem Dev on v on Bicken. Ibidem, 32, 673-6 75. 1883. Beitrag zur Kenntnis der Pl acodermen des n orddautschen Ob erdevons. Abh. K. G es. W i ss. G ot tingen , 30, 1-40. Gott inge n. 1890. Hat Coccos teus vorder e Ruderorgane? N. .i b. Min. etc., 2, p. 1911 . Le ipz ig. 1895. Ube r einige F ischr este d es Norddeu tschen und b ohmischen Devons . Abh. K. G es. W iss. G ot t inge1L, 40, 2, 1-37. Gottingen . KULrZYCKI J . HJ56. On the parasphenoid of th e Brachythoraci (0 parasfenoidzic ryb Brachythoraci). Acta Pa!aeont. PaL, 1, 2, 103-111. Warszawa, LEHMAN J. P. 1956. us Arthrodires du Devonian super ieur du Tafilal et. Emv · Cheri!. Min. Prod. Ind. etc. Notes et Mem., 129, 1-70. Rabat. 'LERICHE M. ,1931. Les poissons famenniens de la Belgique. Mem. Acad. Ro y. Belg.• 10, 1-72. Bruxelles. JULIAN KULCZYC~

LIU H. T . 1945. Kiangyosteus, a new Ar thr odir an fish fr om Szechuan. Ch ina . Acta Palaeont. Sinica, 3, 4, 271-274. P eking. MARKE. J . 1953. Vidy Holonema iz srednego devona Estonskoj SSR. Jubil. Sborn . Ob se. Est est v . Prir. A.N. ESSR. TaBin . NEWBERRY J . S . 1857. Ncw fossil fishes fr om th e Devonian ro cks of Ohio. Amer. J . Sci., 2, 24, 147-159. New H av en . 1868. On som e r em arkable fossil fis hes discovered by Rev . H. Herzer in the black shale (Devonian) at Delaware, Oh io.- I bi dem , 2, 34, 73-78. 1875a. The structure a nd r elations of Dinichthys. Col umbus, 1-64. 1875b. De scription of fos sil fish es. Rept. Ge ol. Su r v . Ohio, 2, 2, 1-40. 1883. The r el ations of Dinichthys, as shown by complete crania r ecently di scov ered by Mr. J ay Terrell in the Huorn sh ale of Ohio. Trans. N. Y. Acad. s ei., 3, 1-20. New York. 1885. Description of some gigantic placoderm fishes recently discovered in the Devoni an of Ohio. I bi dem , 5, 2'5-28 . 1887. De scription of new species of Titanichthys. I bi dem , 6, 164-165. 1888. On th e fossil fishes of the Erie shale of Ohio. I bi dem , 7, 178-180. 1889. The palaeozoic fish es of North America. Monogr . U. S. Geol. Surv., 16. Wash ington. NIKO LSKIYG . V. 1950. Castnaja ichtiologija. Gas. I zdat. , 1-436. Moskwa . OBRUCEV tOBRuTSCHEVI'1 D. W. 1927. Angarichthys, neue Arthrodiren -Gattung aus darn Devon des Flusses Bac h ta (Slblrten). Bull. Com. Geol., 45. 679-688. Le ni ngr ad. ,EI30. Znacenie ieh tiofauny dla str atigrafii sever o-z apadnogo devona. I zv . Gl. Geol.-Razv. U'pravl., 49, 9, 92-99. 1931. tiber Coceosteu s trautscholdi (Eastman). Trav. Mus. GeoL A cad. Sci. URSS, 8, 285-310. Moskva. 1932. Holonemida e d es .r uss is ch en Devons. Trav. Inst. Palcont. Acad. Sci. URSS., 2, 97-115. Mos kva . 1934. Uber zwei Coccosteus-Funde in der Sow jet-Arktis. Trans. Arct. I n st. , 8, 185-19 0. Leningrad . 1937. Vertebrata aus dem Silur und Devon de >; Ur als . Mat. Centro NalLe. I zd. Ge ol.-Razv. Inst., 2, Leningrad. 194{}. Dev ons k ie r yby Si biri i Srednej Azji. Do k l . •4kad.N auk SSSR, 27, 8•. 889- 892. Mosk va. 1941a. Material y po devonskim r ybam SSSR. Trav . Inst . Paleont. A cad . Sci . URSS, 8, 4, 1-19. Mo skv a. 19410. Devonskie r yby Minusinskogo kraja. Ibi dem , 8, 4, 23-417. OBRUCEVA O. P . 19'54. Rod Plourdosteus (Ar thr odira) iz verchnego dev on a SSSR. D okl. Akad. Nauk SSSR, 96, 5, 105'5-1056 . Mo skwa . _ .1956. Ostatki Dinichthys (Ar throd ir a) iz verchnego devona SSSR. I bidem, 108. 2·, 33.:3-3G'5. OKULITCH V. J. 1944. A new Dinichthys from the Devonian of Manitoba. Trans . Roy . Soc. Canada, 3 ser., 38, 4, 65-69. Ottawa. 0RVIG T. 195'1. Histologic studies of placoderms and fossil elasm obr anchs. Arkiv Zool., 2 ser., 2, 2, 321-454. Stockholm. 1957a. Remarks on the ve rtebrate fauna of the lower u pper Dev on ia n of Esc umin ac Bay etc. Ibidem, 10, 6, 367-426. ,1957b. P aleontological not es. 1: On the structure of the bone tissu e in ·the scale s of certain P al aeon isciformes. I bi dem , 10, 12. 481-49{} . UPPER DEVONIAN FISHES 3fl7

RINGUEBERG E. N. 1884. A new Dinichthys fr om the P ortage group of western New York. A rnz r . J . Sci. , 3, 27, 476-4'78. New Have n. ROMER A. S. 1945. Vertebrate P al eon tology. 1- 627. Ch icago. SCHMIDT H. 1938. Uber As pid.ichthys u nd Anornalichthys . Palaean t. Ztsch. , 20, 3/4, 313-317. Berlin. SMITH B. 1909. On som e dinichthy id ar mor pla tes fr om the Mer ycallus sh al e. A m er. N aturalist, 43, 588-597. New Yor k . STENSI(j E. 1931. Upper Devonian Vertebrates fr om Ea st Greenland. M edd. Gnmland, 86, 1-212. K obenhav n. 1934a. On the h ead s of cert ain Arthrod ire s. K. Sv enska Vet. Akad. Handt., 3 ser., 13, 5, 1-79. Stockholm. W3'4b. On the Placod ermi of the Up per Dev on ian ofE'ast Greenland. 1. and Ar thr odira. Medd. Gr0nland, 97, I, 1- 58. Kobenhavn. 1936.· On the Placod ermi of the Upper Devonian of East Greenland, Supple­ ment to part I. Ibidzm , 97, 2, 1-52'. 19'39. On the Placodermi of the Upper Devonian of Ea st Greenland. Second supplement to part 1. Ibidem , 97, 3. J942. On the snout of Arthrodires. K. Sv enska Vet. Acad. Handt., 3 ser., 20, 3, 1-32. Stockholm. 1944. Contribution to the knowledge of the vertebrata... II: Notes on two Arthrodires from the downtonian of Podoli a . Arkiv Zoo!., 35 A, 9, 1-83. Uppsala. 1947. The sensory lines and dermal bones of the cheek in fishes and am phibians . K. Sv enska Ve t . Acad. Hand!., 3 ser ., 24, 3, 1-195. Stockholm: 1948. On the P lacodermi of the Upper Devon ian of East Greenland, II. Palaeozool. Groenland., 2, 1-6,22. Kebenhavn. STETSONH. C. ,1930. Notes on the structure of Dinichthys and Macropatalichthys. Bull. Mus. Compo Zoo!., 71, 19-39. Cambridge, Mass. TRAQUAIR R. H. 1887. Additional notes on the fossil fishes of the Up per Old Red Sandstone of the Moray Area. Proc. Roy. Phys. Soc. Ed inb., 13, 3'76-385. Edinburgh. 1889. Homos tiu s Asmuss, com pa re d with Coccosteus Agassiz. Ibidem, 11, 47-57. 1894. No tes on Palaeozoic fish es. Ann. Ma g. N at . H ist. , 6, 14, 369- 370. London. 1890. On the str uc ture of Coccostsus dec.piens. Ibidem, 6, 5, 125-1 36. L ond on. 1903. T he lower fishes of Gernunden . T r an s. Roy. Soc. Ed inb., 40, 72 3~ 7 3 9 ; Edinburgh. 1909. On fossil r em ains colle cted by J. S. Flett M. A ., D. Sc., from the Old Red. Sandstone of Shetland. Ibidem, 46, 3'21-3 29. TRAUTSCHOLD H. 1880. Uber Dendrodus und Coccosteu s. V erh. Russ. K. Min. Ges., 2, 15, 139-1'56. P etersburg. - 1889. Uber Coccosteus m egalopter yx Trd., Coccosteus obtusus und Ch elio­ phorus verneuili Ag. Ztschr. deutsch. geot. Ges., 41, 35-48. Berlin. WATSON D. M. S. 1932'. On three new spe cies of fi sh from the Old Red Sandstone of Orkney and Shetland. Summ. Pr oqr, Geo!. S urv. Great Brit. 1931, 2, 157,..166. London. 19M. The interpretation of Arthrodires. Proc. Zoo!. Soc. London, 437-464, London. WHITE E. 1952. Australian Arthrodires. Bull. Brit. Mus. (Nat. Hist.), 1, 9, 231-304. London. 368 JULIAN KULCZYCKI

WILLIAMS H. S. 1891. On the p la tes of Ho lone ma rugosa. Pr oc. A mer . Assoc. Adv. Sci., 39, p , 337. WITHLEY G. P . 19i33. New names for fossil fishes. Cop e tn, 1..,146, Ann Arbor . Mi ch ig an. WOODWARD A. S. 189 1. Catalog ue of th e Fossil Fishes in th e British Museu m (Na t. Hist.), Part. 3. London. 1892. F urther contr ibutions to knowl edge of the Canadi an Devonian fi sh es. Ge o!. Mag., 3, 9, 341, 481-485. Lo ndon. 1922. Obser v atio ns on Cro ssopterygian and Arthrodiran fi shes. Proc. L inn . Soc., S ess. 132, 27-36. London .

.JULIAN KULCZYCKI

RYBY GORNO-DEWONSKIE Z GOR SWIF;TOKRZYSKICH (PLACODERMI, ELASMOBRANCHII)

Streszczznie

Przedmiotem pracy sa szcza tk i pl akoderm i spodoustych z utwor6w g6rnego dewonu G6r Swletokrzysklch. Okazy z osad6w Iranskich pochodza z odkrywek W ietrzni, P sich Gorck i Kadzielni w Kielcach, famenskie zas z Kadzielni i Galeztc. Material wypreparowano metoda chemiczna za IXlmOC1j kwasu octowego. W wyniku badan stwierdzono obecnosc nastepujacych nowych form.

RZ1jd Arthrodira Podrzad Brachythoraci

Rodzaj M al erost eu s n. gen.

Ryby 0 szer okiej czaszce z za znaczonym i k antami bocznymi, 0 duzyrn oczo ­ dole ograniczonym tr zema el ementam i kostnymi. Uklad e lemen t6w okolicy policz­ kowej p odobny jak u rodzaju Ph olid ost eus lub B r ac h y osteus. Aparat szczekowy typu mi azdzac ego przypomina szczeki rodzaju D inomylostoma. Od t ego ostatniego roznl sie jed nak sl abym rozwojem p owierzchni mi azdzac e] na infra-gnathale obecnoscla dod atkowego guzka na postero- supra-gn athale.

M al erost eus gorizdroae n. sp. (tekst: fig. 3-6; pl. I, fig. 4-7; pl . II; pl. III, fig . 1-3)

Jest t o jed yny dotychczas znany ga tune k r odza ju Malerost eus , wobec czego charakterystyka gatunku pokrywa sie obe cnie z rodzajowa. Gatunek ten cechuje urzezbleriie, zlozone z g~sto i bezl adnie ro zmieszczonych guzk 6w 0 kopulastym ksztalcie i zmiennej wielkosci. UPPER DEVONIAN FISHES 3619

Materia! stanowi fragment czaszki, obejmujqcy prae-orbitale, sub-orbitale, m ar­ ginale, paranuchale, parasphenoideum i elementy szczekowe oraz luzno znalazione sub-orbitale, post-sub-orbitale i antero-dorso-Iaterale. Wszystkie te okazy pochodza ~ gcrnego franu sasiadujacych ze soba wzgorz Wle trznla i Psie Gorki. Do powyz­ szego gatunku zaliczono tez warunkowo nuchale i postero-mediano-ventrale ze srodkowego franu wzgorza Wietrznia.

Rodzaj Tomaiosteus n. gen.

Ryby 0 szerokiej, plaskiej glowie z niewielkimi oczodolarni, polozonyrni blisko przedniego koiica czaszkl, Pineale Iaczy sie z centraliami i rostale, oddzielajqc w ten sposob calkowicie oba praeorbitalia. Aparat szczekowy typu krajacego, o niezbyt wysokim stopniu sp ecjalizacji. Postero-supra-gnathale posiada wyrostek boczny, czym roznl sie od odpowiednich elernentow innych form 0 tym samym typie aparatu szczekowego.

Tomaiosteus grossi n. sp. (tekst: fig. 7; pl. III, fig. 4, 5)

Jedyny znany gatunek tego rodzaju, poza cechami wymienionymi w diagnozie rodzaju, odznacza sie urzezbieniem zlozonym z rzadko rozsianych, niezbyt duzych guzkow, Material stanowi fragment czaszki obejmujacy pineale, prae-orbitale, centrale i post-supra-gnathale, a pochodzacy ze srodkowego franu wzgorza Wietrznia.

Rodzaj Dinichthys Newberry, 1885 Dinichthys denisoni n. sp. (tekst: fig. 10; pl. VI, fig. 4; pl . VII)

Niewielki przedstawiciel rodzaju Dinichthys, 0 kosciach pozbawionych urzez­ bienia. Mediano-dorsale charakteryzuje sie prostym brzegiem grzebienia, podobnym do wlasciwego D.? jzfjersonensis. Od tej ostatniej formy rozni sill wieksza wyso­ koscia grzebienia i obecnoscia lyzkowatego zaglebienia na wyrostku grzebieniowym. Material stanowia trzy okazy mediano-dorsale, z ktorych jeden pochodzl z warstw 'dolno-cheilocer owych wzgorza Kadzielnia, pozostale zas - z warstw kli­ meniowych miejscowosci Galezice.

Dinichthys ceterus n. sp, (pl. VIII, fig . 1 a-b)

Niewielka forma 0 mediano-dorsale pozbawionym urzezbienia, zaopatrzonyrn

v; niski grzebieii i niema1 poziomo ustawiony wyrostek grzebieniowy. Materia! stanowi okaz mediano-dorsale z warstw dolno-cheilocerowych wzgo­ rza Kadzielnia.

24 Acta Pala eon t olog1ca Polonlca - Vol. II/4 370 JULIAN KULCZYCKI

Rodzaj T itanich t h y s New berr y, 1885 T itani chthys k ozlowskii n. sp. (tekst : fig . ,11; p l. IX, fig. 6)

Stosunkowo niewielka forma , posia d ajaca duze cen tr alia , podobnie jak T . d ar ­ ke i i T. agassi zi. Uk la d k analow linii czu c.owych zapawne t ak i, jak u T. t ermieri . Od T . agassiz i ro zni sie ukszt al towaniem po dwojnego zaglebienia na brzu szne] str onie nuchale. Mat erial stanowia Ira gm enty nuch al e i centrale oraz przypuszczalnego pa ra ­ nuchal e, pochodzaca z w arstw k lim en iowych mi ej scowoscl Gatezice.

Rod za j D ev eon em a n. gen.

Nieduzy przedstawiciel rodziny Holonernidae , 0 zewne trznej powi erzchni kosc i sk or nych pokrytej m al ymi, oddz ielnymi guzkam i, rozsianyrni na ogul bazladnle i jedynie w niektorych m iej scach uklad ajaeymi sie w p odluzne szereg i.

De v eon em a obr ucevi n. sp. (pl. X , fig. 1 a-b)

Jedyny gatunek rodzaju. Definicja gatunkowa pokrywa sie na razie z ro ­ dzajowa, Material stanowi fr agment mediano-dorsale ze sr odkowegc dewonu wzgorzu Wietrznia.

Rodzaj Operchallost eus n. ge n.

Przedstawiciel Brachythorac i, 0 k osc iach sk6rnych pokrytych urzezbteniem, zlozonym z wyraznie odgraniczonych, u podstawy meandrycznych, plaskich zebarek i guzkow. Od Gyroplacosteu s r odzaj t en ro zni sie delikatniejszq ornamentacja, luz­ n ie jszym ulozen iem wynioslcs ci i splaszczeniem ich szczy tu,

Oper cha llosteus via lowi n. sp. (p I. XII, fig . 3)

Definicj a tego jed ynego przed stawi ciel a pokryw a sie z rodzajow a. Material stanow i fragment post er o- ventro- Iateral e ze sr odk ow ego Iranu wzgorza W ietrzn ia. Ponadto wsrcd szcza tkow plak oderm stwier dzono obec nosc nastep ujacyc h for m : 'zna nego z Amery ki Poln, Dinich t hys p ust ul osus Eastman, ga tu nkow nadrenskich A n ornalichthys i 'Tlg ens (Ko enen) i Pac hyosteus bulla J aek el oraz not owanego we franie Rosji H olon ema r ad iat u m (Ob r uc ev) . Bardzo prawdopodobna wydaje sit-; to zsarnosc for my z w ar stw klimeniowych Galezlc , opisanej tu jak o D inichthlls cf. UPPER DEVONIAN FISHES 37\

htbercu!atus, z potnocno-amervkan sk tm D. tuber culatus Newberry i wlqcza nym do tego ostatniego przez autora niniejszej pracy D. belgiclls Leriche z najwyzszego famenu Belgii.

Na podst awie m ateria lu swietokrzyskiego ud a lo sie uzupelnie wia domosci o szczeg61ach budowy niek t6rych z powyzszych form. Przy szczatkach D. p ust ulosus zna leziono nie znane dotychczas u tego 'ga tunk u pa rasphenoid eum, kt6re urnozllwilc wyjasnie n ie przeb iegu arter iae caro tidae in ­ ternae, Dl a Pachyosteus buHa opisano po raz p ierwszy g6r ne ele menty szczekowe, paras phenoideum oraz budow e pancer za brzusznego.

Wsrod szczatko w D. cf. tuberculatus stwierdzono obecnosc infra - gnath ale 0 ty­ powej budowie dla rodzaju D inich t h ys. Element ten u blisk ich nasze] formie ryb arner yk ansk ich i bclgij sk ich ni e byl dotychczas opisa ny . Po za om6wionym i wyzej spotka no szereg ba rd ziej Iragmentarycznych okaz6w, ktorych przynale zno sc nie mogla bye scis le ust alona. Naleza tu szczatki Ox y os te u ~ sp., P!ourdosteus sp. i B othriolepis sp. ze sro dkowego franu Wietrzni oraz St en o­ steus? sp. z famenu Kadzielni. W czesci dodatkowe] opisa no kilka okazow zebo w Cladodus i Dittodus oraz kolec pletwowy w rodza ju Cten acanthus, ja k rown iez dwu nowe kolce pl etwowe: AHenacanthus m alkoUlsk ii n. gen., n. sp . i Sent acanthus zelichowsk ae n . gen., n. sp . charakteryzujqce sie n iem al calkowitym (u pierwszego) lub zupe lnym (u drugiego) brakiern dentyny.

NIEKTORE SZCZEGOLY BUDOW Y BHACHYTHORACI

Postero-supra-gnathalla rodzajow M al erost ell s Tomaiosteus uwidocznia ja rozny ki erunek spa cjalizacji. U pierwszego, podobnie ja k u Dinichthyidae, aparat szczekowy przystosowuie sle do ciecia, u drugiego zas , podobnie ja k u Mylostomi­ dae, do mi azdzenia pozywi enia. Mimo przeciwstawnego kier unku r ozwoj u postero­ supra-gnathalia tych form zachowujq ten sa m zasa dniczy sche mat bu dowy. Swiad­ ezy to, ze ich aparat szczek owy rozw inal sle z jedna kowego typu wyjsciowego. ktorego przyklad em jest chwytny apa rat szczek owy Coccosteidae. Zarowno u Male­ rosteus i Mylostomidae, jak u Tomaios teus i Coccosteidae obecny jest na po stero­ supra-g nathale wyr ostek boc zny. W zwi azku z przystosowan lem do ciec la, postero­ supra-gnathale u Tomaiost eus ulega zwezeniu do cienkiej plytk i z mal o wy dat­ nvrn wyrostkiem bocznym. Ten ostatn i zaniku calkowicie u Dinic hthyidae, kton' pos unely sie da lej w tej specjalizacji. f.T M aLzr olit eus, przeciw nie. w zwi azku z prz ystosowaniem do miazdzcnia pokarmu, postero- su pra-gna thale ulega rozsze­ rzeni u. J ednoczesnie silnie ro zwija sie wyrostek boczny. To samo st wicrdzamy u D inomyl ostoma i M ylost oma. U tego ostatniego ro dzaju, kt6ry osiaga krancowa specjalizacje, postero-s upra-g nathale jest pl askie i niezwykle szerokie, przy cZY rJ1 wyrostek boczny stanowi glow nq mase e lernentu. Na miazdzacej powierzchni zna j - :n2 JULIAN KULCZYCKI dujq sill tu, podobnie jak u Malerostims, dwie wyniostosci. Czy sa one u obu form homologiczne, trudno jest okreslic. Mozliwe, ze mamy tu zjawisko konwer­ gencji. Zestawienie aparatow szczekowych Malerosteus, Dinomylostoma i Mylo­ stoma uwidocznia, ze w procesie przystosowywania sill do miazdzeriia pokarmu po­ szczegolna clementy szczekows ul egaia zmianom w roznym stopniu I nie jedao­ czesnle. Powierzchnia mlazdzaca rozwija sill najpierw na postero-supra-gnathaie, nastepnie na infra-gnathale i w koncu na antaro-supra-gnathala, Stosunek miedzy powierzchniami miazdzacymi poszczegolnych elernentow przypomina stosunek wla­ sciwy innym kregowcom miazdzacyrn pokarm i wywolany jest zapewne tyrni sa­ mymi czynnikami mechaniki aparatu szczekowego, W przeciwienstwie do Dinichthyidae brak wyrostka bocznego na postero­ supra-gnathale u Pachyosteidae wydaje sic: by e pierwotny. Przernawia za tym ogolnie p ierwotny typ aparatu szczek owego u Pachyosteus. Cechami, ktore swlad­ CZq, ze jest on bardziej pierwotny niz u Coccosteidae, sa: brak zroznicowania zab­ kow pod wzgledem postaci i wielkosci oraz bezladne ich rozsianie, b2Z wytworzenla zroznlcowanych szeregow,

Bardzo swoista jest budowa zabkow na szczekach rodzaju Pachyosteus. Majll one [arne miekiszowa i scia nke zlozona z trzech warstw: 1) d eritynowej, 2) przej­ sciowe], z komorkarni ulozonyrni prostopadle do powierzchni,I 3) zewnetrznei, ty­ powej kostnej , z komorkami ulozon ym i rownolegle do powierzchni. Obraz ten przypomina budowe guzkow na ko sciach sk or nych u Holoptychius I pozwala przy­ puszczac, ze: a) guzki na kosciach powlokowych i zabki gebowe sa takimi samymi i po ­ wstalymi w jednakowy sposob tworami; b) u Brachythoraci nastapil proces rozpadu jednolitej powlcki dentynowej na oddzielne zabki i przeksztalcania ich w kostne guzki, podobnie jak u Crossopte­ rygii i Dipnoi;

c) proces ten u Brachythoraci nastapil wczesrnej 1 ostatnia jego faza, w prze­ clw ienstwie do tego, co widzimy u Crossopterygti i Dipnoi, dotknela rowniez zab­ kow gElbowych. Za wystepowaniem wyrnienionego wyzej procesu u Brachythoraci przemawia tez obecnose urzezblenia na kosciach skor nych u form bardziej prymitywnych i zanikanie jego u form posunietych dalej w rozwoiu oraz jednoczesne znikanie dentyny i wreszcie sam jej charakter. Obecnosc dentyny notowana byla u Coccosteus i Plourdosteus, a obecnie stwier­ dzona zostala przez autora rowniez u Pachyosteus. U bardziej wyspecjalizowanych Toma:osteus i Dinichthys dentyny brak. U Malerosteus w infra-gnathale zachowa! siG jeden niewielki uklad dentynowy w przednim zebowatyrn wyrostku. Charakter tego wyrostka swiadczy, ze nigdy nie funkcjonowal jako zab i nalezy go rozpatry­ wac jako twor rudymentarny. Cechy d zntyny u Malerosteus swiadcza, ze tzw. se ­ midentyna (0rvig, 1951) nie stanowi samodzielnego typu dentyny, ale jest mody- UPPER DEVONIAN F ISHES 373 fikacjq wystepujaca na pograniczu miedzy tk anka kostna i tYPOWq dsntyna: cha­ rakterystyczna dl a niej obecnosc jamek komorkowych zwlazana jest z ni sdosk ona­ lym roznicowan iem sill odontoblastow, zale znym od zachodzacego u Brachythoraci pr ocesu zastepowania dentyny przez tkanke ko stna,

Zanik dentyny na szczekach Brachythoraci jest 0 tyle ni ezrozumialy, ze do­ tyczy zarowno form pobieraj qcych pok arm bez ro zdrabniania, jak tsz form 0 tria­ cym i mlazdzac ym typ ie aparatu szczekowego. Zjawi sk o to mozna by tlurnaczyc tym, ze raz rozpoczety proces za nikania den tyny nie rnogl sill zatr zyrnac i b yl ni e­ odwracalny, albo ze dantyn a zastepowana byl a przez inne twor y, spowodowane np . rogowacen iem. Na razie tr udno jest sill wypowledziec za [edna lub druga moz li­ woscia. Wielu autorow podaje obecnosc sla dow scie r ania szczek. Materlal, jak im dy sponuje, nie daj e zdecydowan ych dow odow na korzysc jednej czy drugiej alter­ natywy. Poza swoista budowa zabkow ro dzaj Pach yosteus ciek awy jest rowniez ze wzgledu na budowe parasphenoideum, k tora jest tu, jak sill wydaje, ba r dziej pry­ mitywna niz u D i n ichthy s, a na wet nil u B uchan ost 211s. U Pachyosteus, pod obnie [ak u wiek szoscl Arctolepida, ar teriae carotidae inter nae wnikaja do p odstawy czas zki doboczn ie od parasp henoideum i biegna nastepnie po jego grzbietowej po­ wierzchni. W przeciwi en stwie do tego, co znajdujemy u Din chthy s i szeregu innych rodzaj ow, naczynia te nie tworzyly poprzecznej anastornozy, lecz zlewaly siE: w jedno nieparzys te naczynie, wnikaia ca do jamy czaszkowej. Roznice w budo­ wie parasphen oideum dostarc za z pew nos cia jeszcze jednej wskazowkl dla wyod­ rebnienia poszczegolnych l inii rozwojowych tego szczepu. Uksztaltow anie tylnego brzegu czaszki Brachythoraci swiadczy, ze przy opu­ szczonej glowia odcinek'" mi edz y panewkq stawowa i grani ca m ied zy marginale (lub post-marginal e, 0 He wystepuje) a post-sub-orbital s zachodzil p od pancerz tu­ lowia. Okolicznose ta, wraz z obecnoscla na ws pomnia ny m odci nku ro znych two­ row dl a przyczepu miesn l i wi ezadel, dowodzi, ze szczelina skrza lowa slegal a w gorll co najwyzej do gcrne] granicy post-sub-orbital s. Za ta k im p olozeri.em szczeli ny skrzelowej przem awi a tez fakt, ze u rodzaju Synallchenia nast ap ilo cal­ kowite zros niecie pa ncerza glowowego z tulowiem. Pcn iewa z jest to forma nekto­ niczna, trudno przypuszczac, by nastap ilo u ni ej w tor ne przemieszc zenie szczalin skr zelowych na str one brzuszna i polozenie takie musialo bye wlasciwe wszystk im Brachythoraci. Okaz postero-mediano-ventrale Malerosteus gori zdroae? uwidoczni a ob ecnosc linii przyrostowej, czego na ogo! ni e spo ty ka sie u Brachythoraci. Ob ecnosc jej zwi azana jest z brakiem appozycji war stw k osci na w oln e] powierzch ni plytki, co moze bye wynikiem ucisku wyw ieranego np. przez podloza na cialo tej bentonicz­ nej formy. Przabieg linii przyrostowej swiadczy 0 tyrn, ze zarys poszczegolnych elementow szkieletowych ul egal znacznym zmianom w cia gu r ozwoju osobniczego. Na tym samym ok azie widoczne jest uszkodzenie urazowe w postaci dwoch ubyt­ kow ze sladami regeneracji. Odnowa wyraza sill tu w obacnosci pojedynczych 3.74 J ULIAN K ULCZYCKI

guzkow na dnie uszkodzenia. Podobn e nieciagto na wa rs twianie kosci widz imy na okazie nuchale te jze formy, gdzie powsta je COS w rodzaju "li nii Wes tolla". Wyd aje sie, ze zjawlsko to rna tu charakter patologiczny, chcciaz taki spos6b rosntecia kosci sta je siE: niekiedy regula. U D i n icht h y s d. t uberculat us pow odu je to zapewne roznopostaciowose guzkow urzezbi en ia. Jesl! ch odzi u urzezbienie, to tworzace je guzki moga zw iekszac swe r ozmiary wraz ze wzrostern osobnika, jak to jest u Malerost eus, lub tez miec stala w ielkosc, jak u D in icht h ys pustulosus .

o SYSTEMATYCE BRACHYTHORACI

Budowa parasphenoid eum i charakter zabkow szczekowych rodzaju Pachy ­ osteus swiadcza, ze, zgodnie z pogladarni E. S tenslo, forma ta nie m oze bye wypro­ wadza na z Coccosteidae. Jednoczesnie jednak tr ud no sie zgcdzic, by wyodre bni ona przez Stensi6 gr upa Pachyosteomorpha byla [edno stka na turalna. Zaliczo ne do niej P achyosteidae i Dinichthyidae reprezentujq odl egl e od siebie szczepy.

Ze wzgledu na zbyt szczuple dane 0 budowie Brachythoraci , autor stoi na sta ­ now isk u, ze bardziej celowe jest na razie wyodrebnia nie po szczagoln ych rodzin. bez w iaza nia ich w wyzsze jednos tk i taksonomiczne, tak jak to zn a jdujemy w kl a­ syfikacj i Rom era (1945). Na lezy jedna k zwroclc uwage, ze nie wszystkie wyodreb­ nion e tam r odziny Sq r6wnej wartosci.Tak np .Coccosteidae , w u jeciu Romer.' . o He w og61e sta nowia naturalna jedncstke, to jest on a wyzszego rzedu, niz po­ zosta le rodziny. Autor ni e uwaza tez za usprawiedliwione przeciwstaw ianie Myl o­ stom idae w szys tk im po zostal ym Brachyth oraci, jak to czyni np. L. S. Berg (1955). Osobliwosc ich dotyczy jed ynie aparatu szczekowego, k t6rego uksztaltowan il:' u Brachythoraci jest ni ezwykle roznorodn e.

ROZPRZESTRZE NI ENIE NIEKTORYCH BRACHYTHORAC I

Niewielka Hose odsloniec, z ktorych pochodzi opracowany material, skupieme ich na nieduzyrn odcinku oraz fakt, ze wiekszosc for m wystepuje tylko jako po je­ dy ncze szcza tk i - utrudn iajq wyciaganie w nioskow natury biostr atygraficznej, paleoekologiczne j i pal eogeograficznej. Z pewnoscia jednak nie caly obraz , jak i daja opi san a tu materialy, jest wynikiem przyp adk u.

Na tabeli w tekscie angielskim (p. 352) zestawione sa da ne 0 wystepowanl u po­ szczegolnych fonn w dewonie swietokrzyskim. Widz imy ta m, ze szereg fo nn ogra­ nicza sie do pojedynczych pozi ornow. Sa to jednak przewaznie gatunki, reprczen­ towane w kolekcji przez pojedyncze okazy. Wyjqtek stanowi P ach y osteus bulla i Anomalichthys ingens, Ten ostaini wystepuje tutaj, podobnie jak w Nadrenii, we franie; natom iast Pach yost eus bulla, znany w Nadrenii rowniez we Ir anie, u nas wystepuje w g6rnym famenie. Roznoczasowe wystepowanie jednej formy na r6z- · nych obszarach nie jest zjawiskiem ni ezwyklym. Musimy sie jednak Iiczyc z tym, ze .w danym przypadku rnoz e ono bye wynikiem roznic Iacjalnych. .Mimo to . . ze UPPER DEVONIAN FISHES 375 punk ty, z kt6rych p ocnodza okazy, Ieza bli sk o siebte. osady ich wykazuja dose znaczne rozn ice facjalne. Od sloniecla Wietrzni , Psich Gorak i K adzieln i, ktore do ­ star czyly ok azow fauny franskiej, zn ajdujq sie w obrebie tzw. facj i k iel zck ie i , bar dziej plytkowodnej w porownaniu z facjq ly sogorska, od sl aniajaca sie ni eda ­ leko w Sluchowi cach-Czarnowie. W dolnym i sr odk ow ym franie zaznacza jq s i~ du ze roznice facjalne mi ed zy Kadzieln iq i Wietrzn ia. Podczas gdy na Kadzielni m amy sk aliste wap ien ie rafowe, ktore ni e do st arczyly dotychczas szczatkow ry n, w Wietr zni utwory tego wi eku zlozone sa z bardziej i1astych i bitumicznych wa­ 'p ien i plytowych z Iicznymi szczatk am i ryb. Wsrod ni ch zn a jdu jem y szcreg form (Plourdosteus, Ho lonem a, B othriolepis), ch arakterystycznych r aczej ella u tw or ow przybrzeznych lub ladowych . Chociaz ni e m ozn a w yk luczyc, ze for my te za m ieszk i­ waly otwarte mo rze, Istniej e tez jednak m ozliwcsc, ze szc za tki ich zostaly tu w tor n ie pr zyniesione przez pr ad y; przem awial ab y za tym spor adycznosr ich w y­ stepowania, W gornym franie nast epuje wyr6wnanie facji na da nym - terenie ; zarowno na Kadzielni, jak na Psich Gorkach i Wi etrzni osa dzaja sie plytow e wapienie z Man t ic oceras intumes cens i H ypoVt yridi n a cuboide s. J edno czes n ie nastepuie ujednosta jnien ie fa u ny ryb i wszedzle tu spotyka sie Anomalichthys ingens . Dal sza zmiana fac j i n ast epuje z p oczatkiern fam enu, k iedy na obszarze k ieleck im osad zaly siEi cie nkoplytowe wapienie, m argie i Iupkl glebszego m or za. P oza przejSciowymi dolnym i warstwami famenu na Kadzielni, w osa dach Ia mensk ich okolic Kielc szczatk l r yb sa bardziej ro zpr oszone, Li czniejsze szcza tki w na jw yzszyrn fa me n ie sp otyka sie w Galezrcach (Hi k m od Kielc ), gdz le osadzaly sie wap ienle 0 malej m laz szosci, z boga ta fauna bezkregowc ow, Z ryb, poza zn an ym z dolnego fa menu K ad zielni D in icht h y s deniso ni, wystepuja tu D. cr. t uberculat us, Titanic hthys koz­ zlowskii i Pachy osteus bulla. Forma opisana jako D. cr. tuber cu latus jest prawdo­ po dobnie ide ntyczna z amerykanskirn D. tuberculatus i belgijsk im D. belgicus , ktore autor traktuj e jako jed en gatunek . Za tozsarnoscia tych dwoch ostatnich for m prze­ mawia rn. in. podobiens two facji psammitow z Cond roz w Belgii i facji Catskill w Ameryce P In., gdzie wyste p uj a wspomniane formy obok wsp6lny ch dla obu ob szar6w D i pt er us nelsoni i H ol opt y ch i us murchisoni. D i n ic ht h ys t ub ercu lat us znany jest w centr alnych stanach Ameryki PI n. w utworach sro dkowego dewonu. W czasie odpow iada jqcym franowi forma ta uk azu je sit; rowniez we wschodniej czesci St an ow Zj ednoczon ych A. P. Oba zn aleziska europejskie pochodza z gornego fa me nu. W te n' sposob zarysowuje sle ki erunek ek spa nsji tej formy z zachodu Ameryki P In. przez Belgie do Europy"sr odk owej, Row nie szeroko rozprzestrzeniona Forma okazal sie D. pustu!osus. Poczqtkowo forma ta w srodkowyrn dewonie zajmuje ob szary na zach6d od stanu Kent uck y w Ameryce P In. We franie przenika wraz z Man t i cocer as intumescens do stanow w sch odnich. Jednoczesnie ukazuje sie w Eurr-pie sr odk owej. Trudno .pr zyp uscic. by tak szybko zdolala p ok on ac te wielka pr zestr zen ; ponadto nie jest znana d o- o • •• • • t ychczas w Europie zachodnie j. Wydaje sie wiec bardziej prawdopodobne, ze osro - 376 JUL~ KULCZYCKI

dek jej rozwoju znajdowal sie w zachodniej czesci Ameryki PIn. Iub nawet w p61­ nocnej Azji, i ze stamtad forma ta r ozprzcstrzentla sie w obu k ierunkach, dociera­ ja c [sdnoczesnie do Ameryki wschodniej i Europy srodkowe].

OBJASNIENIA DO ILUSTRACJI

F ig. 1 (p. 286) Rozmieszczenie punkt6w, gdzie zn al eziono szczatk i ryb. Sytuacja gaclogi czn a zaznaczona w uproszczeniu, na podstawie m apy J . Czarnockiego . Skala 1:100.000.

F ig. 2 (p. 2190) Brachythoraci gen. et sp. in det. ( a) ; podluzny przek r6j przez mediano-dorsale. zar ys grzebienia. Fig. 3 (p . 294) Malerosteus gorizdroae n. gen. , n. sp .; sche m atyczna r ek on strukcja szkiel etu z boku; brakujqce czesci kropkow ane. A DL ant ero-dor so-Iater al e, IG infra­ gna thaJe, M marginale, PN para-nuchale, PrO prae-orbitale, PSO po st-sub-orbltale.

F ig. 4 (p. 2:94) Malerosteus gorizdroae n. gen. , n. sp .: tylny brzeg para-nuchale; fg panewka sta­ wowa, pg wyrostek stawo w y, x k ant boczny.

F ig. 5 (p , 296) Malerosteus gorizdroae n, gen., n. sp .; aparat szczek ow y: A elementy g6rne od strony brzusznej , B to samo, od strony boczn ej, C apa rat szczek owy od stron y przysrcdkowej, D-E po stero-supra gnathalia od str ony brzusznej, D Dinomylostoma beech er" E Mylost oma v ar iabil e, ASG antero-supra-gnathal e, IG infra-gnathale; PSG p ostero-supra-gnathale, pI w yrostek boczny postero-supra-gnathale, mt guzak sr odk owy. F ig. 6 (p, 298) Malerosteus gorizdroae?, po stero-mediano-ventrale. Gruba linia ciagta - zarys powierzchni wolnej u osobnika dorostego; cienka linia clagla i przerywana - za­ rys powierzchni wolnej u osobnika mlodego; linia kropkowana - zarys calego ele,­ mentu u osobnika doroslego.

F ig. 7 (p. 299) Tomaiosteus grossi n. gen., n. sp.; pineale: A od strony brzuszne], B od str ony grzbietowej. Fig. 8 (p. 304)

Din'chthys pustulosus; czesciowo zr ekonstruowane (pola kropkowane) czas zki: A ze srodkowego franu, B z dolnego franu Wietrzni; C central e, N nuchale, M marginale, PN para-nuchale, PrO prae-orbitale, pta post-orbitale. Fig. 9 (p. 309) , Dinichthys pustulosus; szkice fragment6w czaszek z Ameryki PIn., wykonane na podstawie fotografii. Oznaczenia - jak fig. 8. UPPER DEVONIAN F ISHES 377

F ig. 10 (p. 314) Dinichthys den;soni n. sp.; przekro] podlu zny przez mediano-dorsale, uwidocz­ niajq cy zarys gr zeb ien ia . Linia przerywanq zaznaczono zarys mniej k ompletnego -okazu. F ig. 11 (p. 320) T itan i cht h y s ko zlowskii n. sp.; schemat tylne] okolicy czaszki. Linia ciagla ­ granice zachowanych fragrnentow: linia przer ywana - przypuszczalny przebieg szw u m ied ~ y nuchal s (N) i cantraliami (C); po le zakropkowane - czesc ok azu o zniszczonej powier zch ni. F ig. 12 (p. 324) Pachy osteus bulla: A r ekonstrukcj a szkielet u z boku; B sche mat przed ni ej czesci pancerza brzu sznego i jego polozenia wzg ledem czaszk i (u gory stron a grz bietowa, u dolu - strona brzuszna), AL antsro-Iaterala, A MV an ter o- mediano­ ve ntrale, C centrale. IL inter o-Iate rale, IG in fra-gn athal e, A \TL an tero -ven tro-Ia­ terale, M m arginal e,N nuchale, PM p ost-marginale, PN p ara-nuchal e, PrO prae­ or b itale, PSO post-sub- orbit al e, PtO po st-orbital s, SO sub-or bitale. Pole zakre­ sk ow ane - przypuszczalne polozenie szczeli ny sk r zalow e].

F ig. 13 (p. 326) Pachyost eu s buLLa, parasphenoideum: A od strony gr zbietowej, B ad strony bocznej, C ad strony brzusznej; fh zaglebienie przysadkowe, m. gr nieparzysty ro­ we k sr odk owy, s. t r rowek poprzeczny.

F ig. 14 (p. 3~0) Pachyosteus bulla: A przekro] poprzeczny przez gor ny brzeg infra-gnathale i zabek ( X 20), B podluzny przekro] zabka ( X 150), C p rzekro] przez scianke zabka (X 300), Cp [ama miekiszowa, D dentyna, 0 tkanka kostna, U warstwa posrednia. d zabek, F ig. 1'5 (p . 3'!3) MaLerosteus gori zdroae n. gen., n. sp ., fragment przekroju poprzecznego przez zebow a ty w yrostek in fra-gna thale (XI50); Cp k anal mieklszowy, 0 tkanka kostna, U wars twa p r zejsciowa , D dentyna. F ig. 16 (p. 344) F ra gm ent przckroju guzka kosci Arctolepida gen. et sp. indet. z famenu K a­ dzielni, uw id oczni aj acy zmiane k sztal tu i ukladu jamek komorkowych w kierunku -ad kanalu naczyniowego ku p owierzch ni zewnetrz ne ] kosci .

PI. I Fig. 1. PLourdost eus sp., m a r gin a le (1. G.); X 3. Fig . 2. P Lourdosteus sp .", infr a-gn athale (1. G .); X 1,5. F ig. 3. B rachythcracl ge n. et sp. in det. ( a), m ediano-ventrale (1. G.) : a po­ wie r zchn ia grz bietowa, b powierzchnia br zuszn a ; wielk. nat. M aler osteus gori zdroae n. g en., n. sp.; w ielk . n at. F ig. 4. a Lewe ante ro -s up ra-gnathale (M. Z.) od strony p rzysrodkowe], b to sarno, z boku, Fig. 5. P rawe a n tero-sup ra-g nathale (M. Z.) od strony przysrodkowei. Fig. 6. a Lewe .p oster o-supr a-gnath ale (M. Z.) od strony bocznej, b to sa rno. -od stro ny przysrodkowe], c to sarno, od str ony brzusznej. Fig. 7. a Lewe infra-gnathale (M. Z.) od str ony gr zbietow ej, b to sarno, z boku. 378 JULIAN KULCZYCKI

PI. II Malerosteus gori zdr oae n. ge n., n. sp .: ;( 0.75 Fig. 1. PrO prae-orbitale, SO su b-or b ita le (lVI. Z.). Fig. 2. Sub-orbltale (1. G.). Fig. 3. Antero-dorso -Iaterale (M.Z.). Fig. 4. PN para-nuchale, M marginale, p ta pos t-orbitale (M. Z.). PI. III Fig. 1. Malerosteus go r iz d roae n. gen., n . sp .; post-sub-orbitale (M. Z.) : X 0.75. Malerosteus gorizdroae?; X 0,75 Fig. 2. a Nuchale (1. G.) od strony brzuszne] , b to sa m e, od strony gr zbie­ towej. Fig; 3. Postero-mediano-ventrale (1. G.). Tomaiosteus grossi n. gen., n . sp , Fig. 4. Fragment czaszki (1. G.) : C centrale, P pineale, PrO prae-orbitale ; X 0,75. Fig. 5. a Postero-supra-gnathale (1. G .) od strony przysrodkowe], wielk. na t. : b to sarno, z boku. PI. IV Dinichthys p ust ulosus Eastman Fig. 1, 2. Fragment czaszki z dolnego franu Wietrzni (M. Z.), ca X 0,33: 1 od str ony grzbietowej, 2 od strony brzusznej; C centr ale, M marginale, N nuchale. PN para-riuchale, pta post-orbitale. PI. V Dinichthys pust ulosus Eastman Fig. 1-'\. Med iano-dorsale (M. Z.), X D,S: 1 od strony brzus zne j, 2 od strouy grzbietowei, 3 z boku. 4 z tylu. PI. VI DinichtiI ys pustulo sus Eastman Fig. 1. Antero-supra-gnathale (M. Z .), wielk . nat. Fig. 2. Fragment cz aszki (M. Z.) z~ srodk owego fran u (M. Z.), ca. X 0,33. Fig. 3. Fragment a ntero-Ia te rale (M. Z.), X 0,75. Fig. 4. D inichthys den iso ni n. sp ., m ediano-dor sale (M. Z.) z war stw ch eil oce ­ rowych K adzielni; X 0,5. Fig. 5. Dinichthys sp., antero- dorso-Iaterale (M. Z.) ; X 0.5. PI. VII Dinichthys denisoni n. sp.

Fig. 1, 2. Dwa mediano-dorsale (1. G.): a od tylu, b z boku, C od str on y br zuszne j; X o.s, PI. VIII Fig. 1. Dinichthys ceterus n. sp ., mediano-dorsale (M. Z.), X 0,5: a z bok u, h od strony brzusznej. Dinichthys cf. tuberculatus Fig. 2. Odcisk infra-gnathale (M. Z.), X 0,33. Fig. 3. Fragment postero-ventro-Iaterale (M. Z.), X 0,7'5. Fig. 4. Oxyosteus sp., mediano-dorsale (1. G.), X 0,75; a od strony grzbietowei, b z przodu. UPPER DEVONIAN FISHES 3'79

PI. IX P ac hyosteus bulla J aekel , F ig. 1. Mediano-dorsale (M. Z.), X 0,5. Fig 2. Antero-laterale (I. G .), X 0,75. Fig. 3. Infr a-gnathale (I. G.), w ie lk. na t. F ig. 4. Fragment czaszki (I. G.), X 0,75. Fig. 5. Fragment infra-gnath ale (I. G.); wielk. nat. Fig. 6. T itanich t hys ko zl ow sk ii n. sp ., nuchal e (M. Z.) od strony br zusznej; X 0,75. PI. X Fig. 1. D ev ean ema oorucev i n. gen., n. sp., mediano-dorsal e (I. G .); a od strony grzbietowej , b od strony- brzusznej, w ielk , nat. Fi g. 2. H olonema radiatum (Rohon in coIl.) Obrucev, a n te r o-dor so-Ia terale (1. G .), X 0,75; a od strony brzusznej, b od strony grzbietowe j. PI. XI Anomalichthys ingens (Koenen) Fig. 1.. Mediano-dorsale (I. G.), od tylu; X 0,5. Fig. 2. To sarno, powierzchnia grzbietowa; X 0,33. Fig. 3. Fragment plytki z charakterystycznq ornamentacjq; w ielk . nat. Fig. 4. Stenosteus? sp., fragment in fra-gnathale (M. Z.); X 3.

PI. XII Fig. 1. Bothriolepis sp , ~1. G .); X 0,75. Fig. 2. Brachythoraci gen. et s p. in det. ( ~), antero-dorso -I aterale (I. G.) X 0,75. Fig 3. Operchallosteus v ia lowi n. ge n., n . sp.: po stero-ventro-laterale (I. G.) ; X 0,75. Fi g. 4. P lacodermi sp. , ni eparzysta ply tka (I. G.) ; X 0,75.

PI. XIII Fig. 1. A lienacan thus malkowskii n . gen ., n . sp . (M.Z .): a z boku, C3. X 0,33; I> od przodu, ca. X 0,33; c fragment kolca z za chowany m i za bkami, X 0,75. F ig. 2. Cten acanth us sp . (M. Z.); w ie lk . n at. Fig. 3. Sentacanthus zelichowsk ae n . ge n., n . sp.; dwie czesci kolca (M. Z.); 7. boku; w ielk . nat. Fig. 4. C ladod us sp . 1 (M. Z.); X 8. Fig. 5. C lad od us sp, 2 (M. Z.); X 4. Fig. 6. Cladodus sp. 3 (M . Z.) ; ca. X 2,5. Fig. 7. Dittodus sp. (M. Z.); X 5. Fig. 8. Dittod us sp, (M. Z.); X 3.

IOflHI\H KYr.btjH~KH PbIBbl H3 BEPXHEfO AEBOHA CBEHTOHPlliUCHI1X fOP Pe310Me

Marepaanoxuacroauiea 'p aooTbl HBR HIO'N:H ocrarxa naamsrpnsrx H TIRClJCT a Ho;IW­ 6ep Hbl X Pbl6 'H3

Cpenx H

Ha OCHI:>BlliHHH ~3Y'HmH oro 'Mil 'repml:Jla

,AHHeHHH aJHIiO 3HaW TeJ1 biHoe H3MClHOlme oxep ­ 'TaHHH x ocretl, 10. POACTBeHHalH CBH3b iMell,AY P achy ost ei dae Ii D i nichth y i dae TaKme ,qaJ1eKa. t\1iK H 1I1elRAY P ach yost eidae If C occosteidae; P achy osteomorpha He llBJ1HIOTCH no acea BepDHTPJOCTH eC'I'eCTOOH!HOH I'PYJflflOi1. 11. D in:chth ys tuberc ula t us H D. belgi cus llBlI HI01'CR nOBHAwMoMy IUpeACT

PI. I

Fig. l. Plourdosteus sp., marginal (1. G.); X 3. Fig. 2. P lourdosteus sp, , infragnathal (I. G.); X 1.5. F ig. 3. Brachythoraci gen. et sp. indet. (a). m edian dorsal (1. G .) (( dorsal v ie w . b ventral view; n at. size.

Malerosteus gorizdroae n. gen., n. sp .; n at. size Fi g.4. a Left a n te r ior supragna t h al (M. Z.), m edial v iew, b same, sid e view . Fig. 5. Right a n ter ior supragn a th al (M. Z.), medial view ; Fig. 6. a Left posterior supragnathal (M . Z.) , side view; b sa me, medial view ; c same, ventral view. Fi g. 7. (( L of't in fragnathal (lV1. Z .), dorsal v iew: b sa m e. sid e v iew .

PI. II

Male r ostens gorizdroae n . gen., n. sp .: X 0.75 Fig. 1. PrO preorbital, SO su bor b ita l (M. Z.). F'ig.2. S uborbital (I. G .). Fig. 3. Anterior dorsolateral (M. Z.). Fig. 4. PN p aranuchal, M m argina l, PtO postorbital (M . Z .).

PI. III Fig. 1. M ale r osteus gorizdroae n. gen., n. sp.; postsuborbital (M. Z .) ; X 0.75. M al erosteus gorizdroae? X 0.75 Fig . 2. a N uchal (1. G.) , ventral v iew; b same. d orsa l v iew. Fig. 3. Posterior median ve n tral (1. G.) . T omai oste ns gr ossi n. gen ., n . sp . Fig. 4. Fragment of the h ead shi eld (I. G.) with C ce ntral, P pi neal. PrO pre ­ orbital; X 0.75. F ig. 5. a P ost erior supragnathal (1. G .), m edia l view, n at. si ze ; b sa me, side vie w .

PI. I V Dinichthys pust ulosus Elastman F ig. 1, 2. Fragment of the head sh ield (lVI. Z .) from the Low er Frasnian of Wietrznia . approx, X 0.33 : 1 dorsal view, 2 ventral view; C central, M margina l. N nuchal. PN paranuchal, PtO postorbital.

PI. V

Dinichthys p ustulosus Ea stman Fig. 1-4. Median do rsal (M . Z .). X 0.5; 1 ve n tral v iew, 2 dorsal v iew, 3 side v ie w . 4 p osterior v ie w.

PI. VI

D inichth us pust utcsus Eastman F ig. l. Anterior sup ragnathal (M. Z.) , nat. si ze. F ig. 2. F ragmen t of th e h ead sh ie ld from the M iddle F'ra sn ia n (M. Z .); approx, X 0.33. Fig. 3. Fragment of the a n ter ior lateral (M.Z .). X 0.75. Fi g. 4. Dinichthys d eni soni n. sp., m edian dorsal (M. Z.) fr om Che iloceras bed ' in Kadzielnia; X 0.5. F ig . 5. D inichthy s sp.. a n terior dorsola teral (M . Z .); >< 0.5. Pi. VII Din ic h thy s denisoni n. sp. F'ig. I , 2. T w o median dorsals (I. G .): I a p ost er ior view, b side view, c ventral view; X 0.5.

PI. VIII Fig. 1. Dinichth ys cet er us n. sp., median dorsal 1M.Z.), X 0,5 ; a side view. b ventral view.

D inich t hys cf, tuber culatus Fi g. 2. Infragn a th al (M. Z .), X 0.33. F ig. 3. Fragmen t of the posterior ve n tr ola te ral (M. Z.), X 0.75. Fig. 4. OXlJosteus sp., median dorsal (I. G.), X 0.75 ; a dorsal view, b front v iew.

PI. IX P achy ost eus bu lla Jaekel Fig. 1. Me dian dorsal (M. Z.), X 0.5. Fig. 2. A nt erior lateral (I. G .), X 0.75. Fig. 3. Infragna th al (I. G .), n at. s ize. Fig. 4. Fragment of the h ead shie ld (I. G.) , X 0.75. F ig. 5. Fragment of the infragnathal (I. G.) , nat. size. Fig. 6. T Ltan ic hth ys k ozl ows k .i n. sp., nuch al (M. Z.), ventral view ; X 0.1:>.

PI. X

Fig. 1. D ev eon zm a obrucevi n. ge n., n . sp., m edian dorsal (I. G. ) ; a dorsal v iew. b ventral view; n at. siz e, Fig. 2. Holon em a radiatum (Rohon in coll.) Ob rucev , anterior do r solateral (I. G.) . X 0.75 ; a ventral vi ew, b dorsal vi ew.

PI. X I Anomalichthys ingens (Koe ne n) F ig. 1. Me di an do rsal (I. G .), postarlor v iew; X 0.5. F ig. 2. S am e, dorsal v iew; X 0.33. Fig. 3. F ragment of the shield pl ate showing or namentation; nat . size. Fig. 4. Sren nsreus? sp., fragment of th e infragn a thal (M. Z.);X 3.

PI. XII

F ig. 1. B othri olepi s sp , (I. G. ),X 0.75. Fig. 2. Brach vthoracl gen . et sp. Ind st, ( ~ ) , a nte rior dor sol ateral (I. G.) X 0.75. Fig. 3. Operchal loste us v i alo w l n. gen., n . sp.; posteriol ventrolateral (I. G.) . X 0.75. Fig. 4. Placod ermi sp., unpaired pl a te (I. G.) ;X 0.75.

Pl. XIII Fig. 1. Alien acan thus mal k owsk ii n. gcn., n. sp , (M. Z.) : a si de v law , approx. X 0.33; b front v iew, approx. X 0.J3; c fragment of the spine showing the sh ap e of denticles, X 0.75. Fig. 2. Ctenacan thus sp. (M. Z.); n at. size. F ig. 3. Sen tacanth us id ch ow sk ae n. gen., n. sp.; two p arts of the sp ine (M. Z .). side view; nat. size. Fig. 4. Cladod us sp . 1 (M. Z.); X 8. F :g. 5. Cladod us sp, 2 1M. Z .); X 4. Fig.n. Cladodus sp, 3 tM. Z.); appr ox . X 2.5. F ig. 7. r. itto du s sp. (M. Z.) ;X 5. Fig. 8. Dittodus sp. 1M. Z.); X 3. ACTA PALAEONTOLOGICA POLONICA.VOL. II J . KULCZYCKI. PL. L

5

7 0

3b

30 .ACTA P ALAEONT OLOGICA PO LONIC A, VOL . II J . K ULCZYCKI. PL. 11

3 ACTA PALAEONTOLOGICA POLONlCA, VOL. II J. KULCZYCKI, PL. III

5 a

2b ACTA PALAEONTOLOGICA POLONICA, VOL. II J . KULCZYCKI, PL. I V ACTA P AL AEONTOLOGICA POLONICA. VOL. II J . KULCZYCKI. PL. V ACTA PALAEONTOLOGICA POLONICA. VOL. II J . KULCZYCKI. PL. VI .ACTA PALAEONTOLOGICA POLONICA. VO L. II J . KULCZYCKI. PL. VII

2 b ACTA PALAEONTOLOGICA POLONICA . VOL. II J. KULCZYCKI. PL. VIIr .ACTA PALAEONTOLOGICA POLONICA , VOL. II J . K ULCZYCKI. PL. IX

3 AC TA PA LAE ONTOLOG IC A POLONICA . VOL. II J . KULCZYCKI. PL. x,

2 0 10

1b L~CT A PALAEONTOLOGICA P OLONICA, VOL. II J . KULCZYCKI, PL. xr

4 ACTA PALAEONTOLOGICA POLONICA, VOL. II J . KULCZYCKI, PL. X U : ACTA PALAEONTOLOGICA POLONICA. VOL. II J . KULCZYCKI. PL. XIII

8

4

3 b 7

10 l b

5 6

2 lc 30